Skip to content

Guides and resources for AI engineers: an agents course, five handbooks, a hands-on second-brain project and a vetted reading list.

277 KB
Raw Blame
/* generated by tools/build_graph_site.py - do not edit */
window.SBOS_GRAPH={"stats":{"pages":130,"sections":22,"words":60110,"terms":57,"tools":107,"resources":114,"links":309,"edges":1236,"nodes":448},"order":["sec:course-0-map","sec:course-1-context","sec:course-2-loop","sec:course-3-gate","sec:course-4-harness","sec:course-5-evals","sec:course-6-production","sec:track-graph","sec:track-jev","sec:track-harness","sec:track-loop","sec:track-evals","sec:01-concepts","sec:02-setup","sec:03-ingestion","sec:04-structuring","sec:05-graphs","sec:06-agents","sec:07-retrieval","sec:08-outputs","sec:09-maintenance","sec:10-troubleshooting"],"nodes":[{"id":"root","type":"root","group":"root","label":"Second Brain OS","desc":"Guides and resources for AI engineers: a course on building agents from a single prompt to production, five handbooks that go deep on one layer each, a hands-on second-brain project, the toolkit, and the links worth your time.","val":34},{"id":"col:course","type":"collection","group":"course","label":"The agents course","desc":"Seven modules from a single prompt to a production agent: theory from the whitepapers, a build in every module. Read in order.","val":14},{"id":"col:handbook","type":"collection","group":"handbook","label":"Handbooks","desc":"Not a path - references. The full menu of techniques, tools and builds for one layer of the course. Dip in anywhere.","val":14},{"id":"col:guide","type":"collection","group":"guide","label":"Second brain","desc":"A hands-on project: a knowledge base an agent builds and maintains for you, in plain markdown. One evening to set up.","val":14},{"id":"col:glossary","type":"collection","group":"term","label":"Glossary","desc":"The ideas the whole site keeps coming back to. Each one links to the pages that lean on it hardest.","val":14},{"id":"col:toolkit","type":"collection","group":"tool","label":"Toolkit","desc":"What you install: agent skills, slash commands, subagents, scripts, the course plugin and the starter vault.","val":14},{"id":"col:resources","type":"collection","group":"res","label":"Resources","desc":"Every vetted link: papers, tools, Obsidian plugins, repositories and reading.","val":14},{"id":"sec:course-0-map","type":"section","group":"course","label":"0 · The map","badge":"C0","desc":"What an agent is, the five layers around the model, and when a workflow beats a loop.","pages":["course-0-map/what-an-agent-is","course-0-map/five-layers","course-0-map/agents-or-workflows"],"val":6.35},{"id":"course-0-map/what-an-agent-is","type":"page","group":"course","label":"What an agent is","sec":"sec:course-0-map","pos":1,"of":3,"path":"docs/course-0-map/what-an-agent-is.md","desc":"Google's Introduction to Agents whitepaper (Kaggle, November 2025) defines an agent as an application that makes plans and takes actions to achieve a goal — not a model that answers a question. The definition sounds bland. The consequence is not: the model is one component of four, and usually not…","words":565,"heads":5,"terms":["term:agent","term:five-layers","term:workflows","term:loop"],"val":2.06},{"id":"course-0-map/five-layers","type":"page","group":"course","label":"Five layers, one system","sec":"sec:course-0-map","pos":2,"of":3,"path":"docs/course-0-map/five-layers.md","desc":"Think of an agent as a new employee on his first day. He is bright, fast, and knows nothing about your company. Everything that decides whether he succeeds is the system you build around him — and that system has exactly five layers.","words":643,"heads":7,"terms":["term:agent","term:five-layers","term:workflows","term:loop","term:gate","term:harness","term:evals"],"val":3.38},{"id":"course-0-map/agents-or-workflows","type":"page","group":"course","label":"Agents or workflows","sec":"sec:course-0-map","pos":3,"of":3,"path":"docs/course-0-map/agents-or-workflows.md","desc":"One decision comes before any building: can you write down the steps? If you can, write them. Code that runs known steps in a known order is a workflow, and a workflow beats an agent on every axis except one.","words":605,"heads":4,"terms":["term:transcripts","term:agent","term:five-layers","term:workflows","term:loop","term:harness","term:evals","term:production"],"val":3.49},{"id":"sec:course-1-context","type":"section","group":"course","label":"1 · Context","badge":"C1","desc":"What the model sees: attention, caching, the four places, sessions and memory.","pages":["course-1-context/how-models-read","course-1-context/the-four-places","course-1-context/context-practice"],"val":6.35},{"id":"course-1-context/how-models-read","type":"page","group":"course","label":"How models read context","sec":"sec:course-1-context","pos":1,"of":3,"path":"docs/course-1-context/how-models-read.md","desc":"A bigger context window looks like free capacity. It is not. The model reads every token you send, but it does not weigh them equally, and past a point each extra token makes the answer worse as well as slower and dearer. This page covers the mechanics; The four places covers the discipline built…","words":585,"heads":4,"terms":["term:lost-in-middle","term:agent","term:context-engineering","term:caching","term:loop"],"val":2.5},{"id":"course-1-context/the-four-places","type":"page","group":"course","label":"The four places","sec":"sec:course-1-context","pos":2,"of":3,"path":"docs/course-1-context/the-four-places.md","desc":"How models read established the physics: attention favours the start and end of the window, every token costs, and caching only pays when the prefix never moves. The discipline that follows is simple to state. Every piece of context belongs in exactly one of four places, and most agent problems…","words":589,"heads":6,"terms":["term:transcripts","term:agent","term:context-engineering","term:caching","term:memory","term:loop"],"val":2.28},{"id":"course-1-context/context-practice","type":"page","group":"course","label":"Context in practice","sec":"sec:course-1-context","pos":3,"of":3,"path":"docs/course-1-context/context-practice.md","desc":"You know the mechanics and the four places. This page is the working session: audit an existing agent, restructure it, and add the two upgrades that matter past a certain size.","words":650,"heads":6,"terms":["term:subagents","term:agent","term:context-engineering","term:caching","term:memory","term:harness"],"val":2.5},{"id":"sec:course-2-loop","type":"section","group":"course","label":"2 · Loop","badge":"C2","desc":"Who decides the next step: goal, checker, stop rule, budget - and the hybrid that survives production.","pages":["course-2-loop/loops-vs-workflows","course-2-loop/the-four-parts","course-2-loop/loop-practice"],"val":6.35},{"id":"course-2-loop/loops-vs-workflows","type":"page","group":"course","label":"Loops versus workflows","sec":"sec:course-2-loop","pos":1,"of":3,"path":"docs/course-2-loop/loops-vs-workflows.md","desc":"A workflow is a sequence you wrote in advance: step one, step two, step three, done. A loop is different in exactly one place — the next step is chosen at run time, by the model, based on what just happened. The cycle is think, act, observe, decide, on repeat: the model proposes an action, the…","words":621,"heads":4,"terms":["term:agent","term:workflows","term:loop","term:stop-conditions","term:production"],"val":2.94},{"id":"course-2-loop/the-four-parts","type":"page","group":"course","label":"The four parts of a loop","sec":"sec:course-2-loop","pos":2,"of":3,"path":"docs/course-2-loop/the-four-parts.md","desc":"Every loop that survives production has the same four parts: a goal with a testable definition of done, a checker that lives outside the model, a stop rule, and a budget counted in both turns and dollars. Remove any one and you have not simplified the loop; you have removed its brakes. This page…","words":628,"heads":5,"terms":["term:lint","term:loop","term:stop-conditions","term:critics","term:llm-judge","term:production"],"val":2.61},{"id":"course-2-loop/loop-practice","type":"page","group":"course","label":"Loop practice: filter, loop, fall back","sec":"sec:course-2-loop","pos":3,"of":3,"path":"docs/course-2-loop/loop-practice.md","desc":"The production pattern is not \"run the agent on everything\". It is filter first, loop briefly, fall back. Atlan's engineering write-up of its on-call agent shows the shape at scale: a deterministic, non-LLM filter suppressed roughly 86% of about 11,000 alerts in a 30-day window before a single…","words":781,"heads":3,"terms":["term:scheduled","term:agent","term:loop","term:stop-conditions","term:critics","term:production"],"val":3.05},{"id":"sec:course-3-gate","type":"section","group":"course","label":"3 · The gate","badge":"C3","desc":"Cheap decisions in front of expensive models: classifiers first, System One models where they earn it.","pages":["course-3-gate/cheap-decisions","course-3-gate/gates-in-practice","course-3-gate/gate-practice"],"val":6.35},{"id":"course-3-gate/cheap-decisions","type":"page","group":"course","label":"Cheap Decisions","sec":"sec:course-3-gate","pos":1,"of":3,"path":"docs/course-3-gate/cheap-decisions.md","desc":"An agent does two kinds of work. It writes — summaries, plans, replies, code — and it decides — is this urgent, which folder does this belong in, does this need a person. Writing is open-ended and genuinely needs a large model. Deciding is bounded: the answer comes from a set you already know. Most…","words":627,"heads":4,"terms":["term:agent","term:gate","term:system-one"],"val":2.39},{"id":"course-3-gate/gates-in-practice","type":"page","group":"course","label":"Gates in Practice","sec":"sec:course-3-gate","pos":2,"of":3,"path":"docs/course-3-gate/gates-in-practice.md","desc":"A gate is only as good as two numbers: how often it is right, and how well it knows when it might be wrong. This page covers both, then the honest current state of the newest tool for the job.","words":683,"heads":4,"terms":["term:gate","term:system-one"],"val":2.17},{"id":"course-3-gate/gate-practice","type":"page","group":"course","label":"Gate Practice","sec":"sec:course-3-gate","pos":3,"of":3,"path":"docs/course-3-gate/gate-practice.md","desc":"The discipline in one sentence: before evaluating any vendor, label a few hundred real examples and train a basic classifier. It takes thirty minutes, and it is the baseline every claim must beat on your data before it earns a place in your pipeline.","words":630,"heads":6,"terms":["term:agent","term:gate","term:system-one","term:llm-judge"],"val":2.39},{"id":"sec:course-4-harness","type":"section","group":"course","label":"4 · Harness","badge":"C4","desc":"The office around the model: containment, guides, sensors, permissions - built from the outside in.","pages":["course-4-harness/the-office","course-4-harness/the-four-rings","course-4-harness/harness-practice"],"val":6.35},{"id":"course-4-harness/the-office","type":"page","group":"course","label":"The Office","sec":"sec:course-4-harness","pos":1,"of":3,"path":"docs/course-4-harness/the-office.md","desc":"Hire the most brilliant engineer alive and put them in a bad office — no keys to the rooms they need, no documentation, no test suite, a manager who rubber-stamps everything — and you get bad work. An agent is a model plus a harness: the model is the hire, the harness is the office. What a harness…","words":576,"heads":5,"terms":["term:agent","term:memory","term:loop","term:harness"],"val":2.61},{"id":"course-4-harness/the-four-rings","type":"page","group":"course","label":"The Four Rings","sec":"sec:course-4-harness","pos":2,"of":3,"path":"docs/course-4-harness/the-four-rings.md","desc":"Build the harness from the outside in: containment, guides, sensors, permissions. The order matters because each ring must hold when every ring inside it fails. A guide can be ignored, a sensor can miss, an approval can be misclicked; a wall does not care. So the outermost ring goes up before the…","words":669,"heads":6,"terms":["term:claude-code","term:mcp","term:hooks","term:git","term:agent","term:harness","term:guardrails"],"val":2.39},{"id":"course-4-harness/harness-practice","type":"page","group":"course","label":"Harness Practice","sec":"sec:course-4-harness","pos":3,"of":3,"path":"docs/course-4-harness/harness-practice.md","desc":"Theory done; now harden a real agent. The worked example is a coding agent on a web app with a Postgres database, run through Claude Code, but every step translates to any harness — including the seventy-line loop from build the loop. Work the rings in order.","words":745,"heads":8,"terms":["term:hooks","term:git","term:transcripts","term:agent","term:loop","term:harness","term:guardrails"],"val":2.94},{"id":"sec:course-5-evals","type":"section","group":"course","label":"5 · Evals","badge":"C5","desc":"The same test every month: behavioural checks on traces, judged judges, golden sets that include failures.","pages":["course-5-evals/two-kinds-of-checks","course-5-evals/judges-and-golden-sets","course-5-evals/evals-practice"],"val":6.35},{"id":"course-5-evals/two-kinds-of-checks","type":"page","group":"course","label":"Two Kinds of Checks","sec":"sec:course-5-evals","pos":1,"of":3,"path":"docs/course-5-evals/two-kinds-of-checks.md","desc":"The previous modules built an agent that plans, calls tools, and edits your vault. This one is about knowing whether it works — before your users tell you. Everything in agent evaluation rests on one distinction: checks on the outcome versus checks on the behaviour. You need both, and they answer…","words":617,"heads":5,"terms":["term:transcripts","term:agent","term:stop-conditions","term:evals","term:llm-judge","term:traces"],"val":3.05},{"id":"course-5-evals/judges-and-golden-sets","type":"page","group":"course","label":"Judges and Golden Sets","sec":"sec:course-5-evals","pos":2,"of":3,"path":"docs/course-5-evals/judges-and-golden-sets.md","desc":"Behavioural checks are code, and code is trustworthy. End-to-end grading usually is not code: \"is this reply correct and grounded\" needs a grader that reads, which in practice means another LLM. A judge is a measurement instrument, and no instrument ships calibrated. This page covers checking the…","words":593,"heads":3,"terms":["term:agent","term:evals","term:golden-set","term:llm-judge","term:production"],"val":2.17},{"id":"course-5-evals/evals-practice","type":"page","group":"course","label":"Evals Practice","sec":"sec:course-5-evals","pos":3,"of":3,"path":"docs/course-5-evals/evals-practice.md","desc":"This page is the minimal build: one afternoon, five steps, one file of cases and one runner, no infrastructure. The eval engineering handbook assembles the full pipeline in three builds; everything you make here carries straight into it.","words":789,"heads":7,"terms":["term:frontmatter","term:scheduled","term:agent","term:evals","term:llm-judge","term:traces"],"val":3.05},{"id":"sec:course-6-production","type":"section","group":"course","label":"6 · Production","badge":"C6","desc":"From demo to deployed: gateways, tracing, cost, security - and the day-one plan across all five layers.","pages":["course-6-production/from-prototype","course-6-production/operating-agents","course-6-production/day-one-plan"],"val":6.35},{"id":"course-6-production/from-prototype","type":"page","group":"course","label":"From Prototype to Production","sec":"sec:course-6-production","pos":1,"of":3,"path":"docs/course-6-production/from-prototype.md","desc":"Every module so far ends with something that works on your machine. This one is about what happens when other people — or a scheduler at 3am — start depending on it. Google's Prototype to Production whitepaper (Kaggle, November 2025) opens with the honest version: \"Building an agent is easy.…","words":609,"heads":4,"terms":["term:scheduled","term:agent","term:caching","term:loop","term:gate","term:evals","term:traces","term:production"],"val":2.61},{"id":"course-6-production/operating-agents","type":"page","group":"course","label":"Operating Agents","sec":"sec:course-6-production","pos":2,"of":3,"path":"docs/course-6-production/operating-agents.md","desc":"Once an agent is deployed, three questions have to be answerable at any moment: what is it doing, what is it costing, and what could it do that you would regret. This page is the operating manual — observability, cost, security, and the loop that governs change.","words":679,"heads":4,"terms":["term:agent","term:caching","term:loop","term:gate","term:harness","term:guardrails","term:tools-tool-use","term:evals","term:traces","term:production"],"val":2.72},{"id":"course-6-production/day-one-plan","type":"page","group":"course","label":"The Day-One Plan","sec":"sec:course-6-production","pos":3,"of":3,"path":"docs/course-6-production/day-one-plan.md","desc":"Six modules is a lot of theory to hold at once, so here is the whole course compressed into one working day. Five layers, and on each one the single cheapest move that pays for itself immediately. None of them needs new infrastructure. All of them can be done before you next let an agent run…","words":704,"heads":6,"terms":["term:agent","term:five-layers","term:loop","term:stop-conditions","term:gate","term:harness","term:guardrails","term:evals"],"val":3.27},{"id":"sec:track-graph","type":"section","group":"handbook","label":"Knowledge graphs","badge":"T1","desc":"Graphs as agent memory: GraphRAG, extraction pipelines, stores — then an evening build of a graph layer over your own vault.","pages":["track-graph/why-graphs","track-graph/graphrag","track-graph/building-graphs-with-llms","track-graph/graph-stores","track-graph/tools","track-graph/build-extract","track-graph/build-query","track-graph/build-use","track-graph/resources"],"val":9.05},{"id":"track-graph/why-graphs","type":"page","group":"handbook","label":"Why Knowledge Graphs","sec":"sec:track-graph","pos":1,"of":9,"path":"docs/track-graph/why-graphs.md","desc":"An agent's memory problem is not storage. It is that facts arrive connected and get stored flat. A knowledge graph keeps the connections: entities as nodes, relationships as typed edges, so that \"what does X have to do with Y\" is a walk rather than a guess.","words":477,"heads":4,"terms":["term:second-brain","term:ingest","term:knowledge-graph","term:typed-links","term:rag","term:vectors","term:agent","term:memory"],"val":3.27},{"id":"track-graph/graphrag","type":"page","group":"handbook","label":"GraphRAG","sec":"sec:track-graph","pos":2,"of":9,"path":"docs/track-graph/graphrag.md","desc":"GraphRAG is retrieval-augmented generation where the index is a graph built from the corpus, not a pile of embedded chunks. The canonical version is Microsoft Research's 2024 paper From Local to Global, which targeted the question vector RAG answers worst: queries about the corpus as a whole.","words":437,"heads":4,"terms":["term:clusters","term:graphrag","term:rag"],"val":2.83},{"id":"track-graph/building-graphs-with-llms","type":"page","group":"handbook","label":"Building Graphs with LLMs","sec":"sec:track-graph","pos":3,"of":9,"path":"docs/track-graph/building-graphs-with-llms.md","desc":"Extraction is the whole game. A graph store will faithfully persist whatever you put in it; whether that is knowledge or noise is decided in the pipeline. This page is the practical shape of that pipeline.","words":431,"heads":5,"terms":["term:agent","term:memory","term:loop"],"val":2.28},{"id":"track-graph/graph-stores","type":"page","group":"handbook","label":"Graph Stores","sec":"sec:track-graph","pos":4,"of":9,"path":"docs/track-graph/graph-stores.md","desc":"Where the graph lives matters less than people think and more than vendors admit. The honest ordering for personal and agent scale runs from plain text upward, not from database downward.","words":432,"heads":5,"terms":["term:wikilink","term:typed-links","term:clusters","term:graphrag","term:agent","term:memory"],"val":3.16},{"id":"track-graph/tools","type":"page","group":"handbook","label":"Tools","sec":"sec:track-graph","pos":5,"of":9,"path":"docs/track-graph/tools.md","desc":"The ecosystem survey, as of September 2026. The field consolidated hard in 2025: the batch-indexing generation went quiet and the incremental, memory-oriented generation took over.","words":381,"heads":5,"terms":["term:vault","term:obsidian","term:knowledge-graph","term:clusters","term:graphrag","term:agent","term:memory"],"val":3.27},{"id":"track-graph/build-extract","type":"page","group":"handbook","label":"Build: Extract the Graph","sec":"sec:track-graph","pos":6,"of":9,"path":"docs/track-graph/build-extract.md","desc":"An evening-sized build in three parts: a queryable graph layer over the vault your agent already maintains. This page gets you from a folder of markdown to two edge files; Build: Query It loads them into NetworkX, and Build: Put It in the Loop hands the queries to the agent.","words":638,"heads":5,"terms":["term:vault","term:claude-code","term:wikilink","term:contradictions","term:rag","term:lost-in-middle","term:agent"],"val":2.72},{"id":"track-graph/build-query","type":"page","group":"handbook","label":"Build: Query It","sec":"sec:track-graph","pos":7,"of":9,"path":"docs/track-graph/build-query.md","desc":"Part two of the build. Build: Extract the Graph left two files at the vault root: edges.csv with the plain links and typededges.csv with the typed pass. This page turns them into answers, in a little over forty lines of Python.","words":581,"heads":3,"terms":["term:vault","term:clusters","term:graphrag","term:rag","term:lost-in-middle"],"val":2.5},{"id":"track-graph/build-use","type":"page","group":"handbook","label":"Build: Put It in the Loop","sec":"sec:track-graph","pos":8,"of":9,"path":"docs/track-graph/build-use.md","desc":"Part three. The graph exists and answers questions; now the agent has to reach it without you in the middle. No MCP server, no framework — the tool interface is \"shell out to a script that prints text\", which Claude Code already does natively and which survives every framework migration.","words":619,"heads":4,"terms":["term:vault","term:ingest","term:orphans","term:hubs","term:clusters","term:lost-in-middle","term:slash-commands","term:lint","term:agent","term:loop"],"val":2.83},{"id":"track-graph/resources","type":"page","group":"handbook","label":"Resources","sec":"sec:track-graph","pos":9,"of":9,"path":"docs/track-graph/resources.md","desc":"Curated, not comprehensive. Everything here was checked to be live in September 2026.","words":385,"heads":4,"terms":["term:knowledge-graph","term:graphrag","term:rag","term:agent","term:memory"],"val":2.94},{"id":"sec:track-jev","type":"section","group":"handbook","label":"Jev engineering","badge":"T2","desc":"The gate is the layer; Jev is one way to build it. Typed decisions with confidence scores instead of generated text — and a build you can run before your Jev access lands.","pages":["track-jev/system-one-models","track-jev/what-jev-is-good-for","track-jev/jev-in-an-agent-stack","track-jev/getting-started","track-jev/build-decision-endpoint","track-jev/build-router","track-jev/build-swap-in-jev","track-jev/resources"],"module":"sec:course-3-gate","val":8.6},{"id":"track-jev/system-one-models","type":"page","group":"handbook","label":"System One Models","sec":"sec:track-jev","pos":1,"of":8,"path":"docs/track-jev/system-one-models.md","desc":"Most AI products today are built on large language models: you send text in, the model writes text back, one token at a time. That is slow and expensive when all your software actually needs is a decision. \"System One models\" are a new category — introduced by TypeSafe AI on 15 September 2026 with…","words":443,"heads":3,"terms":["term:gate","term:system-one"],"val":2.72},{"id":"track-jev/what-jev-is-good-for","type":"page","group":"handbook","label":"What Jev Is Good For","sec":"sec:track-jev","pos":2,"of":8,"path":"docs/track-jev/what-jev-is-good-for.md","desc":"Jev is not a general-purpose model. It answers typed questions about a blob of context, nothing more. That makes the fit question simple: does your task reduce to picking from a bounded answer space? If yes, it is Jev-shaped. If the answer needs to be written, it is not. Background in system one…","words":436,"heads":3,"terms":["term:agent","term:workflows","term:system-one","term:evals"],"val":2.17},{"id":"track-jev/jev-in-an-agent-stack","type":"page","group":"handbook","label":"Jev in an Agent Stack","sec":"sec:track-jev","pos":3,"of":8,"path":"docs/track-jev/jev-in-an-agent-stack.md","desc":"The most credible early use of Jev is not replacing an LLM but sitting next to one. Agent loops are full of small bounded decisions — which model, which tool, is this safe, are we done — and today each one costs a full LLM call. The emerging pattern, a week old and already visible in the LangChain…","words":430,"heads":3,"terms":["term:agent","term:loop","term:gate","term:system-one","term:tools-tool-use"],"val":2.61},{"id":"track-jev/getting-started","type":"page","group":"handbook","label":"Getting Started With Jev","sec":"sec:track-jev","pos":4,"of":8,"path":"docs/track-jev/getting-started.md","desc":"Jev has been publicly documented for about a week. This page reflects what is published as of 21 September 2026; expect it to date quickly.","words":411,"heads":4,"terms":["term:system-one","term:production"],"val":2.39},{"id":"track-jev/build-decision-endpoint","type":"page","group":"handbook","label":"Build: a Typed Decision Endpoint","sec":"sec:track-jev","pos":5,"of":8,"path":"docs/track-jev/build-decision-endpoint.md","desc":"Jev is waitlisted, so most readers cannot call it yet — access status in getting started. What you can build today is the shape it fits into: a function that takes a schema and some state and returns a typed decision with a confidence score. Build that seam now with an ordinary LLM, and the System…","words":543,"heads":4,"terms":["term:gate","term:system-one"],"val":2.06},{"id":"track-jev/build-router","type":"page","group":"handbook","label":"Build: the Confidence-Gated Router","sec":"sec:track-jev","pos":6,"of":8,"path":"docs/track-jev/build-router.md","desc":"The typed decision endpoint earns its keep in front of an agent. The pattern, described in Jev in an agent stack: classify the incoming request cheaply, act directly when the classifier is confident, fall through to the full LLM when it is not. The decision head is the fast path, never the only…","words":501,"heads":3,"terms":["term:agent","term:gate","term:system-one"],"val":2.28},{"id":"track-jev/build-swap-in-jev","type":"page","group":"handbook","label":"Build: the Jev Swap","sec":"sec:track-jev","pos":7,"of":8,"path":"docs/track-jev/build-swap-in-jev.md","desc":"The point of the last two pages was to make this one boring. The decision endpoint and the router agree on one contract — decide(schema, question, state) returning a value and a confidence — so replacing the LLM stand-in with Jev should be a one-file change, not a rewrite. Whether you can make the…","words":524,"heads":3,"terms":["term:gate","term:system-one"],"val":2.06},{"id":"track-jev/resources","type":"page","group":"handbook","label":"Jev Resources","sec":"sec:track-jev","pos":8,"of":8,"path":"docs/track-jev/resources.md","desc":"Jev launched on 15 September 2026. This field is one week old: links below were checked on 21 September 2026 and this page will age fast. Concepts in system one models.","words":313,"heads":4,"terms":["term:agent","term:system-one","term:evals"],"val":2.06},{"id":"sec:track-harness","type":"section","group":"handbook","label":"Agent harnesses","badge":"T3","desc":"The machinery around the model: loops, tools, context engineering, the landscape — and a working harness in an evening, about 150 lines.","pages":["track-harness/what-a-harness-is","track-harness/claude-code-as-harness","track-harness/context-engineering","track-harness/tools-and-mcp","track-harness/harness-landscape","track-harness/build-the-loop","track-harness/build-guardrails","track-harness/build-graduate","track-harness/resources"],"module":"sec:course-4-harness","val":9.05},{"id":"track-harness/what-a-harness-is","type":"page","group":"handbook","label":"What a Harness Is","sec":"sec:track-harness","pos":1,"of":9,"path":"docs/track-harness/what-a-harness-is.md","desc":"A harness is everything around the model that turns text prediction into work done: the loop, the tools, the prompts, the permissions, the plumbing. The model decides; the harness executes, observes, and feeds the result back. Same model, different harness, wildly different agent.","words":467,"heads":7,"terms":["term:context-window","term:agent","term:loop","term:harness","term:guardrails","term:tools-tool-use"],"val":3.27},{"id":"track-harness/claude-code-as-harness","type":"page","group":"handbook","label":"Claude Code as Harness","sec":"sec:track-harness","pos":2,"of":9,"path":"docs/track-harness/claude-code-as-harness.md","desc":"Claude Code is the most instructive harness to study because Anthropic ships the same machinery twice: as a terminal product, and as the Claude Agent SDK — a Python and TypeScript library exposing the identical loop, permission system and context management for your own agents. Read the anatomy…","words":446,"heads":8,"terms":["term:claude-code","term:claude-md","term:mcp","term:skills","term:subagents","term:hooks","term:agent","term:memory","term:harness","term:guardrails"],"val":3.16},{"id":"track-harness/context-engineering","type":"page","group":"handbook","label":"Context Engineering","sec":"sec:track-harness","pos":3,"of":9,"path":"docs/track-harness/context-engineering.md","desc":"Prompt engineering asked \"what words do I use?\". Context engineering asks \"what does the model see at this moment, and why?\". For agents the second question dominates: every tool result, file read and old turn competes for a finite window, and the harness curates it. Anthropic's Effective context…","words":461,"heads":7,"terms":["term:second-brain","term:hooks","term:transcripts","term:agent","term:context-engineering","term:memory","term:stop-conditions","term:harness"],"val":3.38},{"id":"track-harness/tools-and-mcp","type":"page","group":"handbook","label":"Tools and MCP","sec":"sec:track-harness","pos":4,"of":9,"path":"docs/track-harness/tools-and-mcp.md","desc":"Tools are how the model touches the world, and tool design is where most agents quietly fail. A tool is three prompts wearing a schema: the name, the description, and the shape of what comes back. Anthropic's Writing effective tools for agents is the best single guide; this page is the compressed…","words":436,"heads":3,"terms":["term:mcp","term:agent","term:gate"],"val":2.5},{"id":"track-harness/harness-landscape","type":"page","group":"handbook","label":"Harness Landscape","sec":"sec:track-harness","pos":5,"of":9,"path":"docs/track-harness/harness-landscape.md","desc":"A survey of what is alive in September 2026, with one honest take each. They all run the same loop described in what a harness is; they differ in how much machinery they wrap around it and whose opinions you inherit.","words":421,"heads":2,"terms":["term:claude-code","term:agent","term:loop","term:harness","term:guardrails","term:production"],"val":2.5},{"id":"track-harness/build-the-loop","type":"page","group":"handbook","label":"Build: the Loop","sec":"sec:track-harness","pos":6,"of":9,"path":"docs/track-harness/build-the-loop.md","desc":"The anatomy page claims the core of every harness is a while-loop. This page proves it. In roughly seventy lines of Python you get an agent that reads files and runs commands against the live Claude API; the next part adds the machinery that makes it a harness. Budget an evening for all three…","words":547,"heads":4,"terms":["term:agent","term:loop","term:harness","term:tools-tool-use"],"val":2.17},{"id":"track-harness/build-guardrails","type":"page","group":"handbook","label":"Build: Permissions and Context","sec":"sec:track-harness","pos":7,"of":9,"path":"docs/track-harness/build-guardrails.md","desc":"Part one left you with a demo: an agent that runs whatever it likes and grows its transcript without limit. This part adds the three things that make it a harness — a permission gate, a context budget, and a stop condition. The anatomy page called these the parts that let you leave an agent…","words":597,"heads":4,"terms":["term:context-window","term:git","term:transcripts","term:agent","term:stop-conditions","term:gate","term:harness","term:guardrails","term:tools-tool-use"],"val":2.72},{"id":"track-harness/build-graduate","type":"page","group":"handbook","label":"Build: Know When to Graduate","sec":"sec:track-harness","pos":8,"of":9,"path":"docs/track-harness/build-graduate.md","desc":"You now own roughly 150 lines that do what every harness does: loop, execute, gate, budget, stop. The build was the education. This page is the exit interview — what those lines lack, what solves each gap off the shelf, and an honest rule for choosing.","words":554,"heads":5,"terms":["term:claude-code","term:hooks","term:transcripts","term:agent","term:loop","term:stop-conditions","term:gate","term:harness","term:guardrails","term:tools-tool-use"],"val":2.5},{"id":"track-harness/resources","type":"page","group":"handbook","label":"Resources","sec":"sec:track-harness","pos":9,"of":9,"path":"docs/track-harness/resources.md","desc":"Curated, not exhaustive. Everything here was checked in September 2026 and earns its place. Read alongside what a harness is for the vocabulary.","words":516,"heads":4,"terms":["term:mcp","term:transcripts","term:agent","term:context-engineering","term:caching","term:loop","term:harness","term:production"],"val":2.61},{"id":"sec:track-loop","type":"section","group":"handbook","label":"Loop engineering","badge":"T4","desc":"The control system around the agent: stop conditions, critics, context hygiene — and an overnight loop you can trust by morning.","pages":["track-loop/what-loop-engineering-is","track-loop/stop-conditions","track-loop/critics-and-verification","track-loop/context-hygiene","track-loop/patterns","track-loop/build-goal-test","track-loop/build-critic","track-loop/build-overnight","track-loop/resources"],"module":"sec:course-2-loop","val":9.05},{"id":"track-loop/what-loop-engineering-is","type":"page","group":"handbook","label":"What Loop Engineering Is","sec":"sec:track-loop","pos":1,"of":9,"path":"docs/track-loop/what-loop-engineering-is.md","desc":"Loop engineering is the discipline of designing the control system that runs an agent: the thing that decides what to prompt, how to verify the result, whether to retry, and when to stop. You stop doing those jobs by hand, turn by turn, and encode them once. Prompt engineering optimised a single…","words":537,"heads":4,"terms":["term:agent","term:loop","term:stop-conditions","term:critics","term:harness"],"val":2.94},{"id":"track-loop/stop-conditions","type":"page","group":"handbook","label":"Stop Conditions","sec":"sec:track-loop","pos":2,"of":9,"path":"docs/track-loop/stop-conditions.md","desc":"The hardest design problem in a loop is not making the agent go; it is making it stop. An agent will happily report success it has not earned, and a loop with no exit is a bill with no ceiling. Every serious loop carries several independent brakes, any one of which ends the run.","words":545,"heads":5,"terms":["term:agent","term:loop","term:stop-conditions","term:critics"],"val":2.5},{"id":"track-loop/critics-and-verification","type":"page","group":"handbook","label":"Critics and Verification","sec":"sec:track-loop","pos":3,"of":9,"path":"docs/track-loop/critics-and-verification.md","desc":"A loop is only as good as its judge. Generation is now cheap — the model will produce another attempt for pennies — so the bottleneck has moved from writing code to proving it works. Whatever plays the role of critic decides whether your loop compounds progress or compounds garbage.","words":534,"heads":4,"terms":["term:agent","term:loop","term:stop-conditions","term:critics","term:llm-judge"],"val":2.39},{"id":"track-loop/context-hygiene","type":"page","group":"handbook","label":"Context Hygiene","sec":"sec:track-loop","pos":4,"of":9,"path":"docs/track-loop/context-hygiene.md","desc":"The harness handbook covers context engineering: what the model sees within a session, curated turn by turn. Context hygiene is the loop-level version of the same discipline: what survives between turns and between attempts, and — just as important — what gets deliberately destroyed. The harness…","words":501,"heads":4,"terms":["term:second-brain","term:transcripts","term:agent","term:memory","term:loop","term:critics","term:harness"],"val":2.72},{"id":"track-loop/patterns","type":"page","group":"handbook","label":"Patterns","sec":"sec:track-loop","pos":5,"of":9,"path":"docs/track-loop/patterns.md","desc":"Five loop shapes keep reappearing under different names. Each earns its keep on a particular kind of work, and each has failure modes its advocates tend to soft-pedal. They are listed here with both.","words":559,"heads":6,"terms":["term:scheduled","term:loop","term:critics"],"val":2.28},{"id":"track-loop/build-goal-test","type":"page","group":"handbook","label":"Build: the Goal Test","sec":"sec:track-loop","pos":6,"of":9,"path":"docs/track-loop/build-goal-test.md","desc":"Three build pages assemble a working loop around a headless coding agent. This one delivers the skeleton: define done as a script, run the agent, run the goal test, feed failures back, cap the attempts. The examples drive Claude Code's non-interactive mode (claude -p), the reference harness from…","words":527,"heads":3,"terms":["term:agent","term:loop","term:critics","term:harness"],"val":2.72},{"id":"track-loop/build-critic","type":"page","group":"handbook","label":"Build: the Critic","sec":"sec:track-loop","pos":7,"of":9,"path":"docs/track-loop/build-critic.md","desc":"Part one stops on green. This part adds the second judgement: green and clean. A separate critic pass reviews the diff against a narrow rubric, and its defects — not vague advice — feed the next attempt. The theory is in critics and verification; this is the forty extra lines.","words":555,"heads":2,"terms":["term:git","term:critics","term:llm-judge"],"val":2.17},{"id":"track-loop/build-overnight","type":"page","group":"handbook","label":"Build: the Overnight Loop","sec":"sec:track-loop","pos":8,"of":9,"path":"docs/track-loop/build-overnight.md","desc":"The final build runs unattended. Everything from parts one and two still applies; what changes is that nobody is watching, so the rails must make the worst case boring: a budget on iterations, a ratchet that stops the run the moment progress stalls, a morning report in markdown, and git as the…","words":531,"heads":2,"terms":["term:scheduled","term:git","term:agent","term:loop","term:stop-conditions"],"val":2.28},{"id":"track-loop/resources","type":"page","group":"handbook","label":"Resources","sec":"sec:track-loop","pos":9,"of":9,"path":"docs/track-loop/resources.md","desc":"A health warning first: \"loop engineering\" was coined in June 2026 and is three months old as this page is written. The practices underneath are sturdier than the label — Steinberger himself was already asking whether the conversation had moved from loops to graphs by mid-July. Read for the…","words":446,"heads":6,"terms":["term:agent","term:context-engineering","term:loop","term:critics"],"val":2.17},{"id":"sec:track-evals","type":"section","group":"handbook","label":"Eval engineering","badge":"T5","desc":"Measurement as the discipline of AI products: golden sets, judges that do not lie, agent trajectories — and your first suite built in an afternoon.","pages":["track-evals/why-evals","track-evals/designing-evals","track-evals/llm-as-judge","track-evals/agent-evals","track-evals/tooling","track-evals/build-traces","track-evals/build-suite","track-evals/build-ci","track-evals/resources"],"module":"sec:course-5-evals","val":9.05},{"id":"track-evals/why-evals","type":"page","group":"handbook","label":"Why Evals","sec":"sec:track-evals","pos":1,"of":9,"path":"docs/track-evals/why-evals.md","desc":"Evals are the tests of AI products. A prompt change that fixes one complaint quietly breaks three other things; without measurement you find out from users. Teams that ship reliable LLM features treat evaluation as an engineering discipline, not a vibe check before the demo.","words":439,"heads":3,"terms":["term:agent","term:evals","term:traces"],"val":2.83},{"id":"track-evals/designing-evals","type":"page","group":"handbook","label":"Designing Evals","sec":"sec:track-evals","pos":2,"of":9,"path":"docs/track-evals/designing-evals.md","desc":"An eval is only as good as its test cases. The route from \"our bot sometimes messes up\" to a suite you trust runs through failure modes: name what goes wrong, then build cases that provoke it. This page assumes you have done the error analysis described in why evals.","words":456,"heads":5,"terms":["term:agent","term:evals","term:golden-set"],"val":2.39},{"id":"track-evals/llm-as-judge","type":"page","group":"handbook","label":"LLM as Judge","sec":"sec:track-evals","pos":3,"of":9,"path":"docs/track-evals/llm-as-judge.md","desc":"Some checks are code: exact match, regex, JSON schema, \"does the SQL run\". Everything else — tone, faithfulness, whether the answer actually addresses the question — needs a grader that reads. That grader is usually another LLM. Useful, cheap, and untrustworthy until calibrated.","words":415,"heads":4,"terms":["term:evals","term:llm-judge"],"val":2.61},{"id":"track-evals/agent-evals","type":"page","group":"handbook","label":"Agent Evals","sec":"sec:track-evals","pos":4,"of":9,"path":"docs/track-evals/agent-evals.md","desc":"A single completion is a function; an agent is a process. It plans, calls tools, reads results, and recovers (or doesn't) from its own mistakes. Evaluating the final answer alone is like reviewing a chess game by the last move. This page builds on designing evals.","words":476,"heads":5,"terms":["term:agent","term:gate","term:harness","term:evals","term:llm-judge","term:traces"],"val":2.72},{"id":"track-evals/tooling","type":"page","group":"handbook","label":"Eval Tooling","sec":"sec:track-evals","pos":5,"of":9,"path":"docs/track-evals/tooling.md","desc":"The market consolidated around two shapes: eval-first platforms with observability bolted on, and tracing-first platforms with evals bolted on. All of the below are alive and maintained as of late 2026. Pick one and move on; the flywheel from why evals matters far more than the vendor.","words":438,"heads":4,"terms":["term:claude-code","term:skills","term:agent","term:gate","term:evals","term:traces"],"val":2.5},{"id":"track-evals/build-traces","type":"page","group":"handbook","label":"Build: Read Your Traces","sec":"sec:track-evals","pos":6,"of":9,"path":"docs/track-evals/build-traces.md","desc":"First of three build pages: this one gets you from \"the bot sometimes messes up\" to a golden set file on disk. Build the suite makes it runnable; the gate page wires it into CI. An afternoon covers all three. Background, if you skipped it: why evals.","words":633,"heads":4,"terms":["term:clusters","term:evals","term:golden-set","term:traces"],"val":2.28},{"id":"track-evals/build-suite","type":"page","group":"handbook","label":"Build: the Suite","sec":"sec:track-evals","pos":7,"of":9,"path":"docs/track-evals/build-suite.md","desc":"You have golden.jsonl from read your traces. This page turns it into a runnable suite: one plain Python file, no framework. Code assertions grade everything they can; one narrow judge handles the rest, per llm as judge. The example target is a support assistant behind a single prompt — swap…","words":648,"heads":3,"terms":["term:golden-set","term:llm-judge"],"val":2.39},{"id":"track-evals/build-ci","type":"page","group":"handbook","label":"Build: the Gate","sec":"sec:track-evals","pos":8,"of":9,"path":"docs/track-evals/build-ci.md","desc":"A suite that runs only when someone remembers is a suite that stops running. This page wires evalsuite.py from build the suite into GitHub Actions: a subset gates every PR, the full suite runs nightly, and the number lands in a scoreboard committed to the repo. The principles are in the CI section…","words":538,"heads":4,"terms":["term:scheduled","term:git","term:privacy","term:workflows","term:gate","term:evals","term:golden-set","term:llm-judge"],"val":2.83},{"id":"track-evals/resources","type":"page","group":"handbook","label":"Eval Resources","sec":"sec:track-evals","pos":9,"of":9,"path":"docs/track-evals/resources.md","desc":"A short list, deliberately. Everything here earns its place; everything it links to is optional. The concepts they teach map onto why evals and the rest of this handbook.","words":385,"heads":3,"terms":["term:agent","term:evals","term:llm-judge","term:traces"],"val":2.28},{"id":"sec:01-concepts","type":"section","group":"guide","label":"Concepts","badge":"01","desc":"Why an LLM-maintained knowledge base beats the note system you abandoned, and what the pattern actually is.","pages":["01-concepts/what-is-a-second-brain","01-concepts/the-save-for-later-paradox","01-concepts/llm-wiki-pattern","01-concepts/two-layers","01-concepts/why-markdown-and-plain-text","01-concepts/pkm-lineage","01-concepts/wiki-vs-rag","01-concepts/what-good-looks-like"],"val":8.6},{"id":"01-concepts/what-is-a-second-brain","type":"page","group":"guide","label":"What a second brain actually is","sec":"sec:01-concepts","pos":1,"of":8,"path":"docs/01-concepts/what-is-a-second-brain.md","desc":"One local folder of plain text that an agent reads and writes, holding what you have read and understood, that you can ask questions.","words":323,"heads":4,"terms":["term:second-brain"],"val":1.84},{"id":"01-concepts/the-save-for-later-paradox","type":"page","group":"guide","label":"The save-for-later paradox","sec":"sec:01-concepts","pos":2,"of":8,"path":"docs/01-concepts/the-save-for-later-paradox.md","desc":"You screenshot the post. You bookmark the article. You save the video to a playlist named \"watch later\" that has four hundred items in it.","words":301,"heads":4,"terms":["term:llm-wiki"],"val":1.84},{"id":"01-concepts/llm-wiki-pattern","type":"page","group":"guide","label":"The LLM wiki pattern","sec":"sec:01-concepts","pos":3,"of":8,"path":"docs/01-concepts/llm-wiki-pattern.md","desc":"The specific pattern this guide implements, published by Andrej Karpathy as a gist on 2026-04-04.","words":356,"heads":5,"terms":["term:llm-wiki","term:two-layers","term:atomic-pages","term:lint","term:agent"],"val":2.5},{"id":"01-concepts/two-layers","type":"page","group":"guide","label":"The two layers","sec":"sec:01-concepts","pos":4,"of":8,"path":"docs/01-concepts/two-layers.md","desc":"Most second-brain setups fail because they pick one shape and try to make it do two jobs.","words":238,"heads":2,"terms":["term:llm-wiki","term:two-layers","term:ingest"],"val":2.17},{"id":"01-concepts/why-markdown-and-plain-text","type":"page","group":"guide","label":"Why markdown and plain text","sec":"sec:01-concepts","pos":5,"of":8,"path":"docs/01-concepts/why-markdown-and-plain-text.md","desc":"The format looks like a detail. It decides whether the system works.","words":398,"heads":6,"terms":["term:vault","term:wikilink","term:git","term:agent"],"val":2.17},{"id":"01-concepts/pkm-lineage","type":"page","group":"guide","label":"Zettelkasten, PARA, evergreen notes","sec":"sec:01-concepts","pos":6,"of":8,"path":"docs/01-concepts/pkm-lineage.md","desc":"None of this is new. The methods people abandoned were mostly right, and it is worth knowing which parts of them still apply, because the good ideas came with a cost that has now changed.","words":423,"heads":6,"terms":["term:atomic-pages","term:agent"],"val":2.06},{"id":"01-concepts/wiki-vs-rag","type":"page","group":"guide","label":"Wiki vs RAG","sec":"sec:01-concepts","pos":7,"of":8,"path":"docs/01-concepts/wiki-vs-rag.md","desc":"The default advice for asking questions of your own documents is to embed them in a vector database. This guide does something else, and the reason is structural rather than a preference about tools.","words":370,"heads":5,"terms":["term:ingest","term:rag","term:vectors"],"val":2.28},{"id":"01-concepts/what-good-looks-like","type":"page","group":"guide","label":"What good looks like","sec":"sec:01-concepts","pos":8,"of":8,"path":"docs/01-concepts/what-good-looks-like.md","desc":"Concrete criteria, so you can tell whether yours is working rather than guessing.","words":355,"heads":5,"terms":["term:vault"],"val":2.06},{"id":"sec:02-setup","type":"section","group":"guide","label":"Setup","badge":"02","desc":"From zero to a vault your agent maintains, in one evening. Follow the pages in order the first time. Nine steps, roughly twenty minutes of typing and however long the interview in step four takes you.","pages":["02-setup/obsidian-install-and-vault","02-setup/claude-code-setup","02-setup/mcp-obsidian","02-setup/claude-md","02-setup/vault-structure","02-setup/project-scoping","02-setup/live-data","02-setup/obsidian-plugins","02-setup/git-and-sync"],"val":9.05},{"id":"02-setup/obsidian-install-and-vault","type":"page","group":"guide","label":"Obsidian and your first vault","sec":"sec:02-setup","pos":1,"of":9,"path":"docs/02-setup/obsidian-install-and-vault.md","desc":"Obsidian is the storage half of the system. It is free, it keeps everything as plain text files on your own machine, and it renders the links between notes as a graph. Nothing here is stored in a company's cloud, and the files stay readable in any text editor if you ever walk away from the app.","words":365,"heads":5,"terms":["term:vault","term:obsidian","term:agent"],"val":2.17},{"id":"02-setup/claude-code-setup","type":"page","group":"guide","label":"Claude Code in the vault","sec":"sec:02-setup","pos":2,"of":9,"path":"docs/02-setup/claude-code-setup.md","desc":"Claude Code is the half that reads and writes. Two ways to run it, and the choice only affects where you type, not what happens.","words":497,"heads":8,"terms":["term:vault","term:obsidian","term:claude-code","term:claude-md","term:mcp","term:ingest","term:index-log","term:scheduled","term:agent","term:guardrails"],"val":3.16},{"id":"02-setup/mcp-obsidian","type":"page","group":"guide","label":"MCP for Obsidian","sec":"sec:02-setup","pos":3,"of":9,"path":"docs/02-setup/mcp-obsidian.md","desc":"MCP is the standard way Claude talks to other applications. For a vault, it adds a door into Obsidian itself rather than into the folder underneath, which means the agent can work with the vault from any session, not only one started inside that directory.","words":514,"heads":8,"terms":["term:vault","term:obsidian","term:mcp","term:clusters","term:agent"],"val":2.39},{"id":"02-setup/claude-md","type":"page","group":"guide","label":"Writing your CLAUDE.md","sec":"sec:02-setup","pos":4,"of":9,"path":"docs/02-setup/claude-md.md","desc":"CLAUDE.md at the vault root is the file the agent reads at the start of every session. It is the difference between an assistant that knows your situation and one you brief from scratch every time.","words":434,"heads":6,"terms":["term:vault","term:claude-md","term:agent"],"val":2.17},{"id":"02-setup/vault-structure","type":"page","group":"guide","label":"Vault structure","sec":"sec:02-setup","pos":5,"of":9,"path":"docs/02-setup/vault-structure.md","desc":"The vault does two different jobs, and they want different shapes. Keeping them in one folder tree while keeping the shapes separate is what makes the system hold up past the first month.","words":418,"heads":4,"terms":["term:vault","term:claude-md","term:raw","term:two-layers","term:index-log","term:agent"],"val":2.39},{"id":"02-setup/project-scoping","type":"page","group":"guide","label":"Project scoping","sec":"sec:02-setup","pos":6,"of":9,"path":"docs/02-setup/project-scoping.md","desc":"This is the step that most setups skip and then wonder why output quality drops as the vault grows.","words":294,"heads":4,"terms":["term:vault","term:claude-md","term:agent"],"val":1.95},{"id":"02-setup/live-data","type":"page","group":"guide","label":"Live data connectors","sec":"sec:02-setup","pos":7,"of":9,"path":"docs/02-setup/live-data.md","desc":"Static notes are half a brain. The other half is what is changing: your calendar, your inbox, the channels where decisions actually get made.","words":298,"heads":5,"terms":["term:vault","term:mcp","term:agent","term:guardrails"],"val":1.95},{"id":"02-setup/obsidian-plugins","type":"page","group":"guide","label":"Plugins worth installing","sec":"sec:02-setup","pos":8,"of":9,"path":"docs/02-setup/obsidian-plugins.md","desc":"Obsidian works fine with none of these. Add them when you hit the specific problem each one solves.","words":283,"heads":6,"terms":["term:vault","term:obsidian","term:frontmatter","term:git","term:agent"],"val":3.05},{"id":"02-setup/git-and-sync","type":"page","group":"guide","label":"Git, sync and backups","sec":"sec:02-setup","pos":9,"of":9,"path":"docs/02-setup/git-and-sync.md","desc":"A vault is a folder of text files, which makes it close to an ideal git repository, and git is what makes an agent safe to run unattended.","words":331,"heads":5,"terms":["term:vault","term:obsidian","term:ingest","term:git","term:privacy","term:agent"],"val":2.39},{"id":"sec:03-ingestion","type":"section","group":"guide","label":"Ingestion","badge":"03","desc":"Getting everything you read, watch and write into raw/ with as little friction as possible, then into the wiki without flooding it.","pages":["03-ingestion/web-clipper","03-ingestion/youtube-transcripts","03-ingestion/pdfs-and-books","03-ingestion/chat-exports","03-ingestion/voice-and-meetings","03-ingestion/newsletters-and-email","03-ingestion/bulk-backfill"],"val":8.15},{"id":"03-ingestion/web-clipper","type":"page","group":"guide","label":"Web articles","sec":"sec:03-ingestion","pos":1,"of":7,"path":"docs/03-ingestion/web-clipper.md","desc":"Articles are the bulk of what most people save, and the browser is where the decision to save happens. Anything that requires switching windows loses.","words":413,"heads":6,"terms":["term:obsidian","term:raw","term:ingest","term:slash-commands","term:web-clipper","term:agent"],"val":2.39},{"id":"03-ingestion/youtube-transcripts","type":"page","group":"guide","label":"YouTube and podcasts","sec":"sec:03-ingestion","pos":2,"of":7,"path":"docs/03-ingestion/youtube-transcripts.md","desc":"Video is the highest-volume, lowest-retention source most people have. An hour of a conference talk leaves almost nothing behind a week later. Transcripts fix that, and a transcript is just text, which is exactly what the vault wants.","words":435,"heads":7,"terms":["term:ingest","term:transcripts","term:agent"],"val":2.06},{"id":"03-ingestion/pdfs-and-books","type":"page","group":"guide","label":"PDFs, papers and books","sec":"sec:03-ingestion","pos":3,"of":7,"path":"docs/03-ingestion/pdfs-and-books.md","desc":"PDFs are the format most likely to arrive in your vault as garbage. Two columns become interleaved lines, tables collapse, footnotes land mid-sentence, and the agent quietly builds pages on top of the mess.","words":408,"heads":6,"terms":["term:ingest","term:page-types","term:agent"],"val":2.06},{"id":"03-ingestion/chat-exports","type":"page","group":"guide","label":"Chat exports","sec":"sec:03-ingestion","pos":4,"of":7,"path":"docs/03-ingestion/chat-exports.md","desc":"Your chat history is the most underrated source you own and the most sensitive. Years of problems you worked through, decisions you reasoned about, and explanations tailored to what you did not understand at the time.","words":479,"heads":6,"terms":["term:vault","term:raw","term:ingest","term:privacy","term:agent"],"val":2.5},{"id":"03-ingestion/voice-and-meetings","type":"page","group":"guide","label":"Voice notes and meetings","sec":"sec:03-ingestion","pos":5,"of":7,"path":"docs/03-ingestion/voice-and-meetings.md","desc":"The ideas worth capturing rarely arrive at your desk. Voice is the only capture method that works while walking, driving or standing in a queue, and it is the one most vaults never wire up.","words":423,"heads":6,"terms":["term:vault","term:two-layers","term:ingest","term:page-types","term:transcripts"],"val":2.17},{"id":"03-ingestion/newsletters-and-email","type":"page","group":"guide","label":"Newsletters and email","sec":"sec:03-ingestion","pos":6,"of":7,"path":"docs/03-ingestion/newsletters-and-email.md","desc":"Newsletters are the easiest source to automate and the easiest to over-ingest. Automate the capture, then be strict about what gets a wiki page, because a feed that files itself will happily fill your vault with a hundred pages a month that you never asked for.","words":408,"heads":5,"terms":["term:vault","term:raw","term:ingest","term:page-types","term:web-clipper"],"val":2.17},{"id":"03-ingestion/bulk-backfill","type":"page","group":"guide","label":"Backfilling years of material","sec":"sec:03-ingestion","pos":7,"of":7,"path":"docs/03-ingestion/bulk-backfill.md","desc":"At some point you will want to import everything: the bookmark folder, the downloads directory, three years of chat history, the read-later queue with eight hundred unread items.","words":569,"heads":8,"terms":["term:vault","term:ingest","term:frontmatter","term:index-log","term:slash-commands","term:lint","term:web-clipper","term:agent"],"val":2.72},{"id":"sec:04-structuring","type":"section","group":"guide","label":"Structuring","badge":"04","desc":"The rules that decide whether your vault becomes a graph or a landfill.","pages":["04-structuring/atomic-pages","04-structuring/page-types","04-structuring/linking-rules","04-structuring/frontmatter-schema","04-structuring/naming-and-aliases","04-structuring/dedupe-and-merge","04-structuring/contradictions-and-supersession","04-structuring/index-and-log"],"val":8.6},{"id":"04-structuring/atomic-pages","type":"page","group":"guide","label":"One page, one idea","sec":"sec:04-structuring","pos":1,"of":8,"path":"docs/04-structuring/atomic-pages.md","desc":"A page holds one idea, explained well enough that someone who has never seen the source understands it.","words":351,"heads":5,"terms":["term:atomic-pages","term:page-types","term:agent"],"val":1.84},{"id":"04-structuring/page-types","type":"page","group":"guide","label":"The four page types","sec":"sec:04-structuring","pos":2,"of":8,"path":"docs/04-structuring/page-types.md","desc":"Four types, each with a job. Every page declares its type in frontmatter, which is what lets you query the vault by shape rather than reading everything.","words":394,"heads":6,"terms":["term:vault","term:page-types","term:agent"],"val":1.95},{"id":"04-structuring/linking-rules","type":"page","group":"guide","label":"Linking rules","sec":"sec:04-structuring","pos":3,"of":8,"path":"docs/04-structuring/linking-rules.md","desc":"Links are the entire value of the structure. Without them you have a folder of files with good filenames, which is a search index, and search indexes already exist.","words":475,"heads":8,"terms":["term:vault","term:backlinks","term:hubs"],"val":2.17},{"id":"04-structuring/frontmatter-schema","type":"page","group":"guide","label":"Frontmatter schema","sec":"sec:04-structuring","pos":4,"of":8,"path":"docs/04-structuring/frontmatter-schema.md","desc":"Frontmatter is what makes the vault queryable by something other than text search. It costs nothing to write and it is what Dataview, scripts and the agent itself use to find things by shape.","words":380,"heads":6,"terms":["term:vault","term:page-types","term:frontmatter","term:aliases","term:agent"],"val":2.06},{"id":"04-structuring/naming-and-aliases","type":"page","group":"guide","label":"Naming and aliases","sec":"sec:04-structuring","pos":5,"of":8,"path":"docs/04-structuring/naming-and-aliases.md","desc":"Wikilinks resolve by name, so the name is not cosmetic. It is the key the whole graph is built on.","words":364,"heads":5,"terms":["term:vault","term:mcp","term:backlinks","term:aliases","term:rag","term:agent"],"val":2.39},{"id":"04-structuring/dedupe-and-merge","type":"page","group":"guide","label":"Deduplication and merging","sec":"sec:04-structuring","pos":6,"of":8,"path":"docs/04-structuring/dedupe-and-merge.md","desc":"Duplicates are inevitable. Two sources use different words for the same idea, an alias was missing, a bulk import ran before the vault had the concept. The question is not how to prevent them entirely but how to catch them before they both accumulate links.","words":379,"heads":5,"terms":["term:vault","term:backlinks","term:aliases","term:contradictions","term:git","term:agent"],"val":2.17},{"id":"04-structuring/contradictions-and-supersession","type":"page","group":"guide","label":"Contradictions and supersession","sec":"sec:04-structuring","pos":7,"of":8,"path":"docs/04-structuring/contradictions-and-supersession.md","desc":"This is the rule that separates a second brain from a search engine, and the one most implementations get wrong.","words":448,"heads":6,"terms":["term:vault","term:ingest","term:contradictions","term:index-log"],"val":2.28},{"id":"04-structuring/index-and-log","type":"page","group":"guide","label":"index.md and log.md","sec":"sec:04-structuring","pos":8,"of":8,"path":"docs/04-structuring/index-and-log.md","desc":"Two files that are not knowledge, but without which the knowledge is hard to reach and impossible to audit.","words":447,"heads":6,"terms":["term:vault","term:index-log","term:agent"],"val":2.06},{"id":"sec:05-graphs","type":"section","group":"guide","label":"Graphs","badge":"05","desc":"The part most second-brain guides skip. They tell you to open the graph view, say it looks like a brain, and move on.","pages":["05-graphs/graph-basics","05-graphs/obsidian-graph-view","05-graphs/typed-links","05-graphs/graph-vs-vectors","05-graphs/graphrag","05-graphs/exporting-your-graph","05-graphs/metrics"],"val":8.15},{"id":"05-graphs/graph-basics","type":"page","group":"guide","label":"Graph basics for a knowledge base","sec":"sec:05-graphs","pos":1,"of":7,"path":"docs/05-graphs/graph-basics.md","desc":"Your vault is a graph whether you plan it or not. Pages are nodes, wikilinks are edges. Every structural decision in the previous section was a decision about what shape that graph takes.","words":489,"heads":4,"terms":["term:vault","term:wikilink","term:knowledge-graph","term:hubs","term:clusters","term:agent"],"val":2.17},{"id":"05-graphs/obsidian-graph-view","type":"page","group":"guide","label":"Reading the Obsidian graph view","sec":"sec:05-graphs","pos":2,"of":7,"path":"docs/05-graphs/obsidian-graph-view.md","desc":"The global graph is the screenshot everyone posts. It is the least useful view in the app.","words":374,"heads":5,"terms":["term:obsidian","term:knowledge-graph","term:typed-links","term:orphans","term:hubs","term:clusters"],"val":2.5},{"id":"05-graphs/typed-links","type":"page","group":"guide","label":"Typed links","sec":"sec:05-graphs","pos":3,"of":7,"path":"docs/05-graphs/typed-links.md","desc":"A plain [[wikilink]] says two pages are related. It does not say how, and at scale that is a real loss.","words":399,"heads":6,"terms":["term:ingest","term:wikilink","term:frontmatter","term:contradictions","term:typed-links","term:agent"],"val":2.28},{"id":"05-graphs/graph-vs-vectors","type":"page","group":"guide","label":"Graph vs embeddings","sec":"sec:05-graphs","pos":4,"of":7,"path":"docs/05-graphs/graph-vs-vectors.md","desc":"The usual advice for asking questions of your own documents is to embed them and retrieve by similarity. This guide does not, and the reason is worth understanding rather than taking on faith.","words":454,"heads":6,"terms":["term:vault","term:atomic-pages","term:vectors"],"val":2.17},{"id":"05-graphs/graphrag","type":"page","group":"guide","label":"GraphRAG and where it fits","sec":"sec:05-graphs","pos":5,"of":7,"path":"docs/05-graphs/graphrag.md","desc":"GraphRAG is the approach of building a graph from a corpus first, then answering questions against the graph rather than against raw chunks. Microsoft Research published the best-known version, \"From Local to Global: A Graph RAG Approach to Query-Focused Summarization\".","words":415,"heads":5,"terms":["term:vault","term:knowledge-graph","term:clusters","term:graphrag"],"val":2.17},{"id":"05-graphs/exporting-your-graph","type":"page","group":"guide","label":"Exporting the graph","sec":"sec:05-graphs","pos":6,"of":7,"path":"docs/05-graphs/exporting-your-graph.md","desc":"Obsidian's graph view is a viewer. To compute anything, you need the graph outside it.","words":385,"heads":7,"terms":["term:vault","term:obsidian","term:knowledge-graph","term:hubs"],"val":2.28},{"id":"05-graphs/metrics","type":"page","group":"guide","label":"Graph metrics worth tracking","sec":"sec:05-graphs","pos":7,"of":7,"path":"docs/05-graphs/metrics.md","desc":"Four numbers, checked monthly. Their direction over time tells you whether the system is working, which is not obvious from using it day to day.","words":391,"heads":7,"terms":["term:vault","term:ingest","term:backlinks","term:orphans"],"val":2.61},{"id":"sec:06-agents","type":"section","group":"guide","label":"Agents","badge":"06","desc":"Turning a one-off setup into a system that maintains itself.","pages":["06-agents/agent-roles","06-agents/scheduled-maintenance","06-agents/subagents","06-agents/hooks","06-agents/skills-and-commands","06-agents/safety-and-guardrails"],"val":7.7},{"id":"06-agents/agent-roles","type":"page","group":"guide","label":"Agent roles","sec":"sec:06-agents","pos":1,"of":6,"path":"docs/06-agents/agent-roles.md","desc":"One agent doing everything produces mediocre work at every job. The instructions that make a good ingestor are not the ones that make a good reviewer, and stuffed into one prompt they dilute each other.","words":502,"heads":9,"terms":["term:vault","term:ingest","term:hubs","term:agent"],"val":2.39},{"id":"06-agents/scheduled-maintenance","type":"page","group":"guide","label":"Scheduling","sec":"sec:06-agents","pos":2,"of":6,"path":"docs/06-agents/scheduled-maintenance.md","desc":"Scheduled runs are what make the vault feel alive. You wake up and it has filed what you clipped, noticed a contradiction, and left you three lines about what changed.","words":445,"heads":6,"terms":["term:vault","term:raw","term:ingest","term:index-log","term:scheduled","term:git","term:web-clipper"],"val":2.83},{"id":"06-agents/subagents","type":"page","group":"guide","label":"Subagents and parallel work","sec":"sec:06-agents","pos":3,"of":6,"path":"docs/06-agents/subagents.md","desc":"Subagents run with their own context. That is the whole point: a subagent reading thirty pages to answer a question does not fill the main session's context with those thirty pages, only with the answer.","words":323,"heads":5,"terms":["term:ingest","term:index-log","term:subagents","term:agent"],"val":2.06},{"id":"06-agents/hooks","type":"page","group":"guide","label":"Hooks","sec":"sec:06-agents","pos":4,"of":6,"path":"docs/06-agents/hooks.md","desc":"Hooks fire on events rather than on a schedule. They are the right tool for things that must happen every time, which is exactly the set of things a prompt instruction will eventually miss.","words":320,"heads":5,"terms":["term:ingest","term:hooks","term:scheduled","term:git"],"val":1.95},{"id":"06-agents/skills-and-commands","type":"page","group":"guide","label":"Skills and slash commands","sec":"sec:06-agents","pos":5,"of":6,"path":"docs/06-agents/skills-and-commands.md","desc":"Two mechanisms that look similar and do different jobs.","words":410,"heads":5,"terms":["term:vault","term:ingest","term:contradictions","term:skills","term:slash-commands","term:scheduled","term:agent","term:workflows"],"val":2.61},{"id":"06-agents/safety-and-guardrails","type":"page","group":"guide","label":"Guardrails","sec":"sec:06-agents","pos":6,"of":6,"path":"docs/06-agents/safety-and-guardrails.md","desc":"The rule this whole section rests on:","words":476,"heads":7,"terms":["term:vault","term:ingest","term:scheduled","term:git","term:privacy","term:agent","term:guardrails"],"val":2.61},{"id":"sec:07-retrieval","type":"section","group":"guide","label":"Retrieval","badge":"07","desc":"How to actually ask your vault things and get answers worth trusting.","pages":["07-retrieval/asking-questions","07-retrieval/query-patterns","07-retrieval/search-tools","07-retrieval/rag-on-top","07-retrieval/context-budget"],"val":7.25},{"id":"07-retrieval/asking-questions","type":"page","group":"guide","label":"Asking the vault","sec":"sec:07-retrieval","pos":1,"of":5,"path":"docs/07-retrieval/asking-questions.md","desc":"You built the vault so answers would come from what you actually read. An agent that quietly falls back on its own knowledge destroys that, and it does so invisibly: the answer is fluent, plausible, and unverifiable.","words":358,"heads":6,"terms":["term:vault","term:index-log","term:agent"],"val":1.95},{"id":"07-retrieval/query-patterns","type":"page","group":"guide","label":"Query patterns","sec":"sec:07-retrieval","pos":2,"of":5,"path":"docs/07-retrieval/query-patterns.md","desc":"Some questions a vault answers far better than a search engine or a general model. These are the ones worth building the habit around.","words":391,"heads":8,"terms":["term:vault","term:contradictions"],"val":2.06},{"id":"07-retrieval/search-tools","type":"page","group":"guide","label":"Search tools","sec":"sec:07-retrieval","pos":3,"of":5,"path":"docs/07-retrieval/search-tools.md","desc":"Retrieval in a plain-text vault is mostly grep. That sounds primitive and is almost always enough.","words":298,"heads":6,"terms":["term:obsidian","term:agent"],"val":1.95},{"id":"07-retrieval/rag-on-top","type":"page","group":"guide","label":"Adding RAG when you need it","sec":"sec:07-retrieval","pos":4,"of":5,"path":"docs/07-retrieval/rag-on-top.md","desc":"This guide argues against a vector database for a personal vault. That argument has a threshold, and it is worth naming rather than treating as doctrine.","words":433,"heads":6,"terms":["term:vault","term:ingest","term:aliases","term:rag","term:vectors","term:context-window","term:stop-conditions"],"val":2.5},{"id":"07-retrieval/context-budget","type":"page","group":"guide","label":"Context budget","sec":"sec:07-retrieval","pos":5,"of":5,"path":"docs/07-retrieval/context-budget.md","desc":"Every page the agent reads costs context, and quality degrades well before the window is full. Reading more is not reading better.","words":399,"heads":6,"terms":["term:vault","term:context-window","term:subagents","term:agent","term:stop-conditions"],"val":2.39},{"id":"sec:08-outputs","type":"section","group":"guide","label":"Outputs","badge":"08","desc":"The vault earns its keep when things come out of it.","pages":["08-outputs/writing-from-the-vault","08-outputs/research-reports","08-outputs/publishing-and-export","08-outputs/teaching-yourself"],"val":6.8},{"id":"08-outputs/writing-from-the-vault","type":"page","group":"guide","label":"Writing from the vault","sec":"sec:08-outputs","pos":1,"of":4,"path":"docs/08-outputs/writing-from-the-vault.md","desc":"Writing is where the vault stops being an input and starts being leverage. The research is already done, cited, and connected. What is left is the argument.","words":333,"heads":5,"terms":["term:vault","term:page-types","term:loop"],"val":1.84},{"id":"08-outputs/research-reports","type":"page","group":"guide","label":"Research reports","sec":"sec:08-outputs","pos":2,"of":4,"path":"docs/08-outputs/research-reports.md","desc":"A research report is what you produce when a question needs more than the vault currently holds. The vault is the starting point, not the whole answer.","words":272,"heads":5,"terms":["term:vault","term:ingest"],"val":1.84},{"id":"08-outputs/publishing-and-export","type":"page","group":"guide","label":"Publishing and export","sec":"sec:08-outputs","pos":3,"of":4,"path":"docs/08-outputs/publishing-and-export.md","desc":"Some of your vault is worth showing to other people. Most of it is not, and the mechanics of that split are worth setting up before you publish anything.","words":332,"heads":6,"terms":["term:vault","term:obsidian"],"val":1.84},{"id":"08-outputs/teaching-yourself","type":"page","group":"guide","label":"Learning from your own vault","sec":"sec:08-outputs","pos":4,"of":4,"path":"docs/08-outputs/teaching-yourself.md","desc":"Reading something is not learning it. A vault full of well-written pages you never revisit is a library, and libraries do not make anyone knowledgeable.","words":354,"heads":6,"terms":["term:vault"],"val":1.84},{"id":"sec:09-maintenance","type":"section","group":"guide","label":"Maintenance","badge":"09","desc":"What keeps the thing alive after the novelty wears off.","pages":["09-maintenance/lint-and-health","09-maintenance/review-cadence","09-maintenance/versioning-with-git","09-maintenance/backups-and-portability","09-maintenance/privacy-and-secrets","09-maintenance/scaling"],"val":7.7},{"id":"09-maintenance/lint-and-health","type":"page","group":"guide","label":"Linting the vault","sec":"sec:09-maintenance","pos":1,"of":6,"path":"docs/09-maintenance/lint-and-health.md","desc":"Structural rot is silent. Nothing errors. The vault just slowly stops answering well because a third of its pages are unreachable and nobody noticed.","words":294,"heads":5,"terms":["term:vault","term:ingest","term:index-log","term:lint","term:git"],"val":2.39},{"id":"09-maintenance/review-cadence","type":"page","group":"guide","label":"Review cadence","sec":"sec:09-maintenance","pos":2,"of":6,"path":"docs/09-maintenance/review-cadence.md","desc":"Lint keeps the structure sound. Review is what keeps you in the loop, and it is the only maintenance task that produces something you read rather than something that happens.","words":312,"heads":6,"terms":["term:vault","term:contradictions"],"val":2.17},{"id":"09-maintenance/versioning-with-git","type":"page","group":"guide","label":"Versioning","sec":"sec:09-maintenance","pos":3,"of":6,"path":"docs/09-maintenance/versioning-with-git.md","desc":"Git is what makes an agent safe to run unattended. Without it, a bad run is a mystery; with it, every change is a diff you can read and revert in one command.","words":328,"heads":6,"terms":["term:vault","term:obsidian","term:git","term:agent"],"val":2.06},{"id":"09-maintenance/backups-and-portability","type":"page","group":"guide","label":"Backups and portability","sec":"sec:09-maintenance","pos":4,"of":6,"path":"docs/09-maintenance/backups-and-portability.md","desc":"This is the whole page, and it is the mistake almost everyone makes.","words":328,"heads":6,"terms":["term:vault","term:obsidian","term:wikilink","term:git"],"val":1.95},{"id":"09-maintenance/privacy-and-secrets","type":"page","group":"guide","label":"Privacy","sec":"sec:09-maintenance","pos":5,"of":6,"path":"docs/09-maintenance/privacy-and-secrets.md","desc":"A mature second brain is one of the highest-signal documents about a person that exists. Everything you read, worked on, worried about and asked, connected and indexed.","words":389,"heads":7,"terms":["term:vault","term:ingest","term:privacy"],"val":2.39},{"id":"09-maintenance/scaling","type":"page","group":"guide","label":"Scaling past a thousand pages","sec":"sec:09-maintenance","pos":6,"of":6,"path":"docs/09-maintenance/scaling.md","desc":"Most of this guide assumes a vault of a few hundred to a few thousand pages, which is where personal vaults live for years. Past that, some things strain.","words":429,"heads":7,"terms":["term:vault"],"val":1.84},{"id":"sec:10-troubleshooting","type":"section","group":"guide","label":"Troubleshooting","badge":"10","desc":"Concrete failures, with fixes.","pages":["10-troubleshooting/common-failures","10-troubleshooting/agent-writes-garbage","10-troubleshooting/broken-links","10-troubleshooting/vault-too-big","10-troubleshooting/faq"],"val":7.25},{"id":"10-troubleshooting/common-failures","type":"page","group":"guide","label":"The failures everyone hits","sec":"sec:10-troubleshooting","pos":1,"of":5,"path":"docs/10-troubleshooting/common-failures.md","desc":"Almost always capture, not the system. Clipping takes three clicks instead of one, or the phone has no path into the vault, so material stops arriving and there is nothing to process.","words":331,"heads":7,"terms":["term:vault","term:claude-md","term:ingest","term:hubs","term:agent"],"val":2.5},{"id":"10-troubleshooting/agent-writes-garbage","type":"page","group":"guide","label":"The agent writes garbage","sec":"sec:10-troubleshooting","pos":2,"of":5,"path":"docs/10-troubleshooting/agent-writes-garbage.md","desc":"Pages that are technically about the right thing and useless to read.","words":320,"heads":6,"terms":["term:claude-md","term:page-types","term:transcripts","term:agent"],"val":2.28},{"id":"10-troubleshooting/broken-links","type":"page","group":"guide","label":"Broken links and orphans","sec":"sec:10-troubleshooting","pos":3,"of":5,"path":"docs/10-troubleshooting/broken-links.md","desc":"Obsidian updates links when you rename inside the app. An agent renaming a file directly does not, and that is the most common cause of broken links in an agent-maintained vault.","words":322,"heads":6,"terms":["term:vault","term:ingest","term:aliases","term:index-log","term:orphans","term:lint","term:agent"],"val":2.5},{"id":"10-troubleshooting/vault-too-big","type":"page","group":"guide","label":"The vault got too big","sec":"sec:10-troubleshooting","pos":4,"of":5,"path":"docs/10-troubleshooting/vault-too-big.md","desc":"Answers get slower and vaguer. The agent reads twenty pages and returns something general. The graph view is a hairball. You cannot remember whether a topic is covered, and checking takes longer than searching the web.","words":316,"heads":8,"terms":["term:vault","term:orphans","term:agent"],"val":2.06},{"id":"10-troubleshooting/faq","type":"page","group":"guide","label":"FAQ","sec":"sec:10-troubleshooting","pos":5,"of":5,"path":"docs/10-troubleshooting/faq.md","desc":"Do I need MCP?","words":439,"heads":0,"terms":["term:vault","term:obsidian","term:claude-code","term:mcp","term:ingest","term:knowledge-graph","term:skills","term:agent","term:memory"],"val":2.72},{"id":"hub:skills","type":"hub","group":"tool","label":"Skills","desc":"One per workflow in the second-brain project. The agent loads a skill only when the task matches it.","inpages":[["04-structuring/index-and-log",3],["06-agents/agent-roles",3],["06-agents/skills-and-commands",3]],"val":4.5},{"id":"hub:commands","type":"hub","group":"tool","label":"Slash commands","desc":"Scoped entry points into the skills, typed as /name.","inpages":[["03-ingestion/bulk-backfill",3],["06-agents/skills-and-commands",3]],"val":4.5},{"id":"hub:agents","type":"hub","group":"tool","label":"Subagents","desc":"Six subagents, four of them read-only by design.","inpages":[["06-agents/agent-roles",3]],"val":4.5},{"id":"hub:scripts","type":"hub","group":"tool","label":"Scripts","desc":"Dependency-free Python for link checking, stats, graph export and chat import.","inpages":[["03-ingestion/chat-exports",3]],"val":4.5},{"id":"hub:plugin","type":"hub","group":"tool","label":"agents-course plugin","desc":"Tools from the agents course: audit your context, write done as a script, bootstrap your eval suite, and a loop critic agent.","url":"https://github.com/undefined-ui/second-brain-os/tree/main/plugins/agents-course","inpages":[["course-1-context/context-practice",3],["course-2-loop/loop-practice",3],["course-3-gate/gate-practice",3],["course-4-harness/harness-practice",3],["course-5-evals/evals-practice",3]],"val":4.5},{"id":"hub:vault","type":"hub","group":"tool","label":"Starter vault","desc":"vault-template/: the wiki structure, the project pipeline, CLAUDE.md and page templates, ready to copy.","url":"https://github.com/undefined-ui/second-brain-os/tree/main/vault-template","inpages":[["02-setup/obsidian-install-and-vault",3],["02-setup/claude-md",3],["02-setup/obsidian-plugins",3]],"val":4.5},{"id":"skill:second-brain-backfill","type":"skill","group":"tool","label":"second-brain-backfill","desc":"Import a large archive into the vault in controlled batches: triage what is worth ingesting, process oldest first, checkpoint after every batch, and keep cost visible. Use this skill whenever the user wants to import years of bookmarks, an export from another notes app, a downloads folder, a chat history dump, or says…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-backfill","val":2.0},{"id":"skill:second-brain-changed-my-mind","type":"skill","group":"tool","label":"second-brain-changed-my-mind","desc":"Trace where the user's recorded position on something has shifted: superseded claims, contradictions between sources of different dates, and what evidence moved them. Use this skill when the user asks what they have changed their mind about, how their thinking on a topic developed, what they used to believe, or wants…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-changed-my-mind","val":2.0},{"id":"skill:second-brain-chat-import","type":"skill","group":"tool","label":"second-brain-chat-import","desc":"Triage an exported chat history into what to ingest, what to archive and what to delete, with a privacy pass first. Use this skill when the user has exported their Claude or ChatGPT history, asks to import old conversations, or asks what in their chat archive is worth keeping. Do NOT use for ingesting a single pasted…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-chat-import","val":2.0},{"id":"skill:second-brain-graph","type":"skill","group":"tool","label":"second-brain-graph","desc":"Analyse the vault's link graph and report on its shape: orphan rate, average degree, components, hubs, bridges and clusters, with what each number means for retrieval. Use this skill when the user asks about the health or structure of their graph, wants metrics, asks which pages are most connected or most isolated, or…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-graph","val":2.0},{"id":"skill:second-brain-ingest","type":"skill","group":"tool","label":"second-brain-ingest","desc":"Turn raw source material in a second-brain vault into linked wiki pages: read the source, split it into concepts and entities, write or update pages, connect them to existing pages, and record the run in the log. Use this skill whenever the user drops a file into raw/, pastes an article, transcript or PDF and asks to…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-ingest","val":2.0},{"id":"skill:second-brain-lint","type":"skill","group":"tool","label":"second-brain-lint","desc":"Audit a second-brain vault for structural problems and repair them: broken wikilinks, orphan pages, empty stubs, frontmatter that violates the schema, near-duplicate pages, and stale index entries. Use this skill whenever the user asks to lint, clean, audit or check the health of their vault, says pages feel messy or…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-lint","inpages":[["04-structuring/dedupe-and-merge",1]],"val":2.0},{"id":"skill:second-brain-merge","type":"skill","group":"tool","label":"second-brain-merge","desc":"Merge two pages that describe the same thing, preserving every distinct claim, redirecting inbound links, and recording the merge. Use this skill when the user approves a duplicate merge, says \"these are the same page\", \"merge X into Y\", or asks to clean up near-duplicates that a lint run proposed. Do NOT use to merge…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-merge","val":2.0},{"id":"skill:second-brain-metrics","type":"skill","group":"tool","label":"second-brain-metrics","desc":"Record a dated snapshot of the vault's health numbers and compare against previous snapshots so trends are visible. Use this skill for a monthly or quarterly check, when the user asks whether the vault is improving, or asks to track its numbers over time. Do NOT use for a one-off graph analysis, which is…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-metrics","val":2.0},{"id":"skill:second-brain-privacy","type":"skill","group":"tool","label":"second-brain-privacy","desc":"Audit the vault for material that should not be in it: credentials, other people's private information, confidential work, and anything the user would not want synced or backed up. Use this skill when the user asks about privacy, is about to share or sync the vault, has just imported chat history or meeting notes, or…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-privacy","val":2.0},{"id":"skill:second-brain-project","type":"skill","group":"tool","label":"second-brain-project","desc":"Set up or update a project in the vault: the four-stage folder pipeline, a project-scoped CLAUDE.md with one goal, and links to the concept pages it depends on. Use this skill when the user starts something new they will work on over time, says \"create a project\", \"set up a folder for X\", or asks to define the goal…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-project","val":2.0},{"id":"skill:second-brain-publish","type":"skill","group":"tool","label":"second-brain-publish","desc":"Prepare a subset of the vault for publishing: check opt-in flags, find private material, and report what would leak through links before anything is exported. Use this skill when the user wants to publish notes as a site or digital garden, asks what is safe to make public, or is setting up an export. Do NOT use for…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-publish","val":2.0},{"id":"skill:second-brain-query","type":"skill","group":"tool","label":"second-brain-query","desc":"Answer a question from a second-brain vault by reading its own pages, citing the pages used, and saying plainly when the vault does not contain the answer. Use this skill whenever the user asks what they know about a topic, what they have read on something, what connects two ideas, what they have changed their mind…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-query","val":2.0},{"id":"skill:second-brain-quiz","type":"skill","group":"tool","label":"second-brain-quiz","desc":"Generate questions from the user's own concept pages so they can find out what they saved but never learned, and compare their answers against what the pages say. Use this skill when the user wants to test themselves, revise, make flashcards from their notes, or asks what they have forgotten. Do NOT use for generic…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-quiz","val":2.0},{"id":"skill:second-brain-rename","type":"skill","group":"tool","label":"second-brain-rename","desc":"Rename a page safely: update the file, fix every inbound wikilink, add the old name as an alias, and log it. Use this skill whenever a page needs a different canonical title, the user says \"rename X to Y\", asks to fix an inconsistent title, or after a lint run flags naming problems. Do NOT use for merging two pages…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-rename","val":2.0},{"id":"skill:second-brain-report","type":"skill","group":"tool","label":"second-brain-report","desc":"Produce a research report from the vault: scope the question against existing pages, name the gaps, and write findings with citations, disagreements and limits. Use this skill when the user asks for a research report, a deep write-up on a topic, or wants to understand a subject thoroughly using what they have…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-report","val":2.0},{"id":"skill:second-brain-review","type":"skill","group":"tool","label":"second-brain-review","desc":"Produce a periodic review of a second-brain vault: what was added, which topics are growing, what contradicts what, which questions are still open, and what is worth reading next. Use this skill whenever the user asks for a weekly or monthly review, asks what changed in their vault, what they have been learning about…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-review","val":2.0},{"id":"skill:second-brain-transcript","type":"skill","group":"tool","label":"second-brain-transcript","desc":"Clean a raw transcript before ingestion: punctuate, paragraph, label speakers, fix mistranscribed technical terms, and split long recordings by topic. Use this skill when the user drops an auto-generated transcript, subtitle file, podcast or lecture transcript, meeting recording text, or asks to process a video or…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-transcript","val":2.0},{"id":"skill:second-brain-write","type":"skill","group":"tool","label":"second-brain-write","desc":"Draft from the vault: outline from concept pages, keep citations to the user's own sources, and surface where their material disagrees. Use this skill when the user wants to write an article, post, essay or newsletter based on what they have collected, or asks what they could write about a topic. Do NOT use for…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/skills/second-brain-write","val":2.0},{"id":"cmd:aliases","type":"command","group":"tool","label":"/aliases","desc":"Find missing aliases","sub":"structuring","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/aliases.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:archive","type":"command","group":"tool","label":"/archive","desc":"Move cold material out of the wiki","sub":"maintenance","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/archive.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:ask","type":"command","group":"tool","label":"/ask","desc":"Ask a question answered only from the vault","sub":"retrieval","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/ask.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:audit","type":"command","group":"tool","label":"/audit","desc":"Audit what an agent did","sub":"safety","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/audit.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:backfill","type":"command","group":"tool","label":"/backfill","desc":"Bulk import an archive in batches","sub":"ingestion","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/backfill.md","via":"skill:second-brain-backfill","inpages":[["03-ingestion/bulk-backfill",1]],"val":0.8},{"id":"cmd:bridges","type":"command","group":"tool","label":"/bridges","desc":"Find the pages holding the graph together","sub":"graph","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/bridges.md","via":"skill:second-brain-graph","val":0.8},{"id":"cmd:changed-my-mind","type":"command","group":"tool","label":"/changed-my-mind","desc":"What I have revised","sub":"retrieval","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/changed-my-mind.md","via":"skill:second-brain-changed-my-mind","val":0.8},{"id":"cmd:claude-md","type":"command","group":"tool","label":"/claude-md","desc":"Build or update your CLAUDE.md","sub":"setup","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/claude-md.md","via":"skill:second-brain-project","val":0.8},{"id":"cmd:clusters","type":"command","group":"tool","label":"/clusters","desc":"Show what the vault is actually about","sub":"graph","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/clusters.md","via":"skill:second-brain-graph","val":0.8},{"id":"cmd:commit","type":"command","group":"tool","label":"/commit","desc":"Commit the current state with a useful message","sub":"maintenance","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/commit.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:commitments","type":"command","group":"tool","label":"/commitments","desc":"List open commitments","sub":"projects","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/commitments.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:compare","type":"command","group":"tool","label":"/compare","desc":"Compare two things from your own sources","sub":"retrieval","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/compare.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:connect","type":"command","group":"tool","label":"/connect","desc":"Find the path between two ideas","sub":"retrieval","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/connect.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:contradictions","type":"command","group":"tool","label":"/contradictions","desc":"List unresolved contradictions","sub":"structuring","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/contradictions.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:contradicts","type":"command","group":"tool","label":"/contradicts","desc":"Argue against me","sub":"retrieval","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/contradicts.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:decisions","type":"command","group":"tool","label":"/decisions","desc":"List decisions made","sub":"projects","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/decisions.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:dedupe","type":"command","group":"tool","label":"/dedupe","desc":"Find near-duplicate pages","sub":"structuring","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/dedupe.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:doctor","type":"command","group":"tool","label":"/doctor","desc":"Check the setup is working","sub":"setup","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/doctor.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:draft","type":"command","group":"tool","label":"/draft","desc":"Draft from the vault","sub":"outputs","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/draft.md","via":"skill:second-brain-write","val":0.8},{"id":"cmd:dry-run","type":"command","group":"tool","label":"/dry-run","desc":"Preview a run without writing","sub":"safety","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/dry-run.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:explain","type":"command","group":"tool","label":"/explain","desc":"Explain it back and find the gaps","sub":"outputs","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/explain.md","via":"skill:second-brain-quiz","val":0.8},{"id":"cmd:export","type":"command","group":"tool","label":"/export","desc":"Export a page or set of pages","sub":"outputs","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/export.md","via":"skill:second-brain-publish","val":0.8},{"id":"cmd:gaps","type":"command","group":"tool","label":"/gaps","desc":"What is missing from my understanding","sub":"retrieval","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/gaps.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:graph-export","type":"command","group":"tool","label":"/graph-export","desc":"Export the graph for outside analysis","sub":"graph","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/graph-export.md","via":"skill:second-brain-graph","val":0.8},{"id":"cmd:graph","type":"command","group":"tool","label":"/graph","desc":"Report the shape of the graph","sub":"graph","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/graph.md","via":"skill:second-brain-graph","val":0.8},{"id":"cmd:handoff","type":"command","group":"tool","label":"/handoff","desc":"Prepare a handoff brief","sub":"projects","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/handoff.md","via":"skill:second-brain-report","val":0.8},{"id":"cmd:health","type":"command","group":"tool","label":"/health","desc":"Quick health check","sub":"maintenance","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/health.md","via":"skill:second-brain-graph","val":0.8},{"id":"cmd:hubs","type":"command","group":"tool","label":"/hubs","desc":"Find pages that swallowed the graph","sub":"graph","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/hubs.md","via":"skill:second-brain-graph","val":0.8},{"id":"cmd:index","type":"command","group":"tool","label":"/index","desc":"Rebuild the index","sub":"structuring","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/index.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:ingest-chats","type":"command","group":"tool","label":"/ingest-chats","desc":"Import an exported chat history","sub":"ingestion","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/ingest-chats.md","via":"skill:second-brain-chat-import","val":0.8},{"id":"cmd:ingest-highlights","type":"command","group":"tool","label":"/ingest-highlights","desc":"Ingest book or article highlights","sub":"ingestion","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/ingest-highlights.md","via":"skill:second-brain-ingest","val":0.8},{"id":"cmd:ingest-mine","type":"command","group":"tool","label":"/ingest-mine","desc":"Ingest your own finished work","sub":"outputs","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/ingest-mine.md","via":"skill:second-brain-ingest","val":0.8},{"id":"cmd:ingest-newsletter","type":"command","group":"tool","label":"/ingest-newsletter","desc":"Ingest newsletters without duplicating","sub":"ingestion","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/ingest-newsletter.md","via":"skill:second-brain-ingest","val":0.8},{"id":"cmd:ingest-paper","type":"command","group":"tool","label":"/ingest-paper","desc":"Ingest an academic paper","sub":"ingestion","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/ingest-paper.md","via":"skill:second-brain-ingest","val":0.8},{"id":"cmd:ingest-pdf","type":"command","group":"tool","label":"/ingest-pdf","desc":"Ingest a PDF","sub":"ingestion","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/ingest-pdf.md","via":"skill:second-brain-ingest","val":0.8},{"id":"cmd:ingest-url","type":"command","group":"tool","label":"/ingest-url","desc":"Clip and ingest a web page","sub":"ingestion","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/ingest-url.md","via":"skill:second-brain-ingest","val":0.8},{"id":"cmd:ingest-voice","type":"command","group":"tool","label":"/ingest-voice","desc":"Ingest a voice note","sub":"ingestion","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/ingest-voice.md","via":"skill:second-brain-transcript","val":0.8},{"id":"cmd:ingest-youtube","type":"command","group":"tool","label":"/ingest-youtube","desc":"Ingest a video or podcast transcript","sub":"ingestion","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/ingest-youtube.md","via":"skill:second-brain-transcript","val":0.8},{"id":"cmd:ingest","type":"command","group":"tool","label":"/ingest","desc":"Ingest new material from raw/ into the wiki","sub":"ingestion","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/ingest.md","via":"skill:second-brain-ingest","inpages":[["03-ingestion/web-clipper",1],["06-agents/skills-and-commands",1]],"val":0.8},{"id":"cmd:init","type":"command","group":"tool","label":"/init","desc":"Scaffold a new vault","sub":"setup","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/init.md","via":"skill:second-brain-project","val":0.8},{"id":"cmd:install","type":"command","group":"tool","label":"/install","desc":"Install the skills, commands and agents","sub":"setup","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/install.md","val":0.8},{"id":"cmd:know","type":"command","group":"tool","label":"/know","desc":"What do I know about a topic","sub":"retrieval","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/know.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:link","type":"command","group":"tool","label":"/link","desc":"Find and add missing connections","sub":"structuring","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/link.md","via":"skill:second-brain-ingest","val":0.8},{"id":"cmd:lint","type":"command","group":"tool","label":"/lint","desc":"Audit structure and repair what is mechanical","sub":"maintenance","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/lint.md","via":"skill:second-brain-lint","inpages":[["03-ingestion/bulk-backfill",1],["09-maintenance/lint-and-health",1]],"val":0.8},{"id":"cmd:merge","type":"command","group":"tool","label":"/merge","desc":"Merge two pages","sub":"structuring","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/merge.md","via":"skill:second-brain-merge","val":0.8},{"id":"cmd:metrics","type":"command","group":"tool","label":"/metrics","desc":"Record a dated metrics snapshot","sub":"maintenance","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/metrics.md","via":"skill:second-brain-metrics","val":0.8},{"id":"cmd:monthly","type":"command","group":"tool","label":"/monthly","desc":"The monthly structural review","sub":"maintenance","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/monthly.md","via":"skill:second-brain-metrics","val":0.8},{"id":"cmd:orphans","type":"command","group":"tool","label":"/orphans","desc":"Find unreachable pages","sub":"graph","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/orphans.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:outline","type":"command","group":"tool","label":"/outline","desc":"Outline a piece from your concept pages","sub":"outputs","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/outline.md","via":"skill:second-brain-write","val":0.8},{"id":"cmd:privacy","type":"command","group":"tool","label":"/privacy","desc":"Audit what should not be in the vault","sub":"safety","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/privacy.md","via":"skill:second-brain-privacy","val":0.8},{"id":"cmd:project-status","type":"command","group":"tool","label":"/project-status","desc":"Where a project stands","sub":"projects","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/project-status.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:project","type":"command","group":"tool","label":"/project","desc":"Create a project","sub":"projects","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/project.md","via":"skill:second-brain-project","val":0.8},{"id":"cmd:prune","type":"command","group":"tool","label":"/prune","desc":"Find what is safe to remove","sub":"maintenance","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/prune.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:publish","type":"command","group":"tool","label":"/publish","desc":"Check what is safe to publish","sub":"outputs","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/publish.md","via":"skill:second-brain-publish","val":0.8},{"id":"cmd:quiz","type":"command","group":"tool","label":"/quiz","desc":"Test yourself on your own pages","sub":"outputs","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/quiz.md","via":"skill:second-brain-quiz","val":0.8},{"id":"cmd:rename","type":"command","group":"tool","label":"/rename","desc":"Rename a page and fix every link","sub":"structuring","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/rename.md","via":"skill:second-brain-rename","val":0.8},{"id":"cmd:report","type":"command","group":"tool","label":"/report","desc":"Write a research report","sub":"outputs","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/report.md","via":"skill:second-brain-report","val":0.8},{"id":"cmd:retype","type":"command","group":"tool","label":"/retype","desc":"Fix a page's type","sub":"structuring","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/retype.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:review","type":"command","group":"tool","label":"/review","desc":"Periodic review of what the vault learned","sub":"maintenance","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/review.md","via":"skill:second-brain-review","inpages":[["09-maintenance/review-cadence",1]],"val":0.8},{"id":"cmd:rollback","type":"command","group":"tool","label":"/rollback","desc":"Undo the last run","sub":"safety","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/rollback.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:schedule","type":"command","group":"tool","label":"/schedule","desc":"Set up scheduled maintenance","sub":"setup","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/schedule.md","via":"skill:second-brain-review","val":0.8},{"id":"cmd:schema","type":"command","group":"tool","label":"/schema","desc":"Check frontmatter against the schema","sub":"structuring","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/schema.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:scope","type":"command","group":"tool","label":"/scope","desc":"Pull knowledge into a project","sub":"projects","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/scope.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:secrets","type":"command","group":"tool","label":"/secrets","desc":"Scan for credentials","sub":"safety","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/secrets.md","via":"skill:second-brain-privacy","val":0.8},{"id":"cmd:sources","type":"command","group":"tool","label":"/sources","desc":"Show what a claim rests on","sub":"retrieval","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/sources.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:split","type":"command","group":"tool","label":"/split","desc":"Split an overloaded page","sub":"structuring","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/split.md","via":"skill:second-brain-ingest","val":0.8},{"id":"cmd:stale","type":"command","group":"tool","label":"/stale","desc":"Find concept pages nobody has touched","sub":"graph","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/stale.md","via":"skill:second-brain-graph","val":0.8},{"id":"cmd:tags","type":"command","group":"tool","label":"/tags","desc":"Audit the tag vocabulary","sub":"structuring","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/tags.md","via":"skill:second-brain-lint","val":0.8},{"id":"cmd:timeline","type":"command","group":"tool","label":"/timeline","desc":"How my sources developed over time","sub":"retrieval","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/timeline.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:trace","type":"command","group":"tool","label":"/trace","desc":"Show which pages an answer used","sub":"retrieval","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/trace.md","via":"skill:second-brain-query","val":0.8},{"id":"cmd:typed-links","type":"command","group":"tool","label":"/typed-links","desc":"Add relation types where they matter","sub":"graph","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/typed-links.md","via":"skill:second-brain-ingest","val":0.8},{"id":"cmd:weekly","type":"command","group":"tool","label":"/weekly","desc":"The weekly review","sub":"maintenance","url":"https://github.com/undefined-ui/second-brain-os/blob/main/commands/weekly.md","via":"skill:second-brain-review","val":0.8},{"id":"agent:curator","type":"agent","group":"tool","label":"curator","desc":"Proposes what to prune, archive or merge. Read-only, proposals never actions. Use quarterly or when the vault has grown past what the user can hold.","tools":"Read, Glob, Grep","url":"https://github.com/undefined-ui/second-brain-os/blob/main/agents/curator.md","inpages":[["06-agents/agent-roles",1]],"val":2.0},{"id":"agent:graph-analyst","type":"agent","group":"tool","label":"graph-analyst","desc":"Analyses the link graph and reports metrics, hubs, bridges and clusters. Read-only. Use for monthly checks or when the vault feels hard to navigate.","tools":"Read, Glob, Grep, Bash","url":"https://github.com/undefined-ui/second-brain-os/blob/main/agents/graph-analyst.md","val":2.0},{"id":"agent:ingestor","type":"agent","group":"tool","label":"ingestor","desc":"Processes new material in raw/ into linked wiki pages. Use when unprocessed sources are waiting.","tools":"Read, Write, Edit, Glob, Grep","url":"https://github.com/undefined-ui/second-brain-os/blob/main/agents/ingestor.md","inpages":[["06-agents/agent-roles",5]],"val":2.0},{"id":"agent:linker","type":"agent","group":"tool","label":"linker","desc":"Finds and adds missing connections between existing wiki pages. Use after bulk imports or on a schedule.","tools":"Read, Edit, Glob, Grep","url":"https://github.com/undefined-ui/second-brain-os/blob/main/agents/linker.md","inpages":[["06-agents/agent-roles",2],["06-agents/scheduled-maintenance",1]],"val":2.0},{"id":"agent:researcher","type":"agent","group":"tool","label":"researcher","desc":"Answers a research question from the vault and flags what is missing. Use for questions that span many pages.","tools":"Read, Glob, Grep","url":"https://github.com/undefined-ui/second-brain-os/blob/main/agents/researcher.md","val":2.0},{"id":"agent:reviewer","type":"agent","group":"tool","label":"reviewer","desc":"Produces the periodic review of what the vault learned. Use weekly or when asked what changed.","tools":"Read, Glob, Grep","url":"https://github.com/undefined-ui/second-brain-os/blob/main/agents/reviewer.md","inpages":[["06-agents/agent-roles",3]],"val":2.0},{"id":"script:chat_export_to_md.py","type":"script","group":"tool","label":"chat_export_to_md.py","desc":"Convert an exported chat history JSON into one markdown file per conversation.","url":"https://github.com/undefined-ui/second-brain-os/blob/main/scripts/chat_export_to_md.py","inpages":[["03-ingestion/chat-exports",1]],"val":1.6},{"id":"script:graph_export.py","type":"script","group":"tool","label":"graph_export.py","desc":"Export a vault's wikilink graph as CSV edges or GraphML.","url":"https://github.com/undefined-ui/second-brain-os/blob/main/scripts/graph_export.py","inpages":[["track-graph/build-extract",2],["track-graph/build-use",2],["05-graphs/exporting-your-graph",2],["01-concepts/why-markdown-and-plain-text",1]],"val":1.6},{"id":"script:link_check.py","type":"script","group":"tool","label":"link_check.py","desc":"Check a second-brain vault for broken wikilinks, orphan pages and stubs.","url":"https://github.com/undefined-ui/second-brain-os/blob/main/scripts/link_check.py","inpages":[["04-structuring/naming-and-aliases",1],["07-retrieval/rag-on-top",1],["09-maintenance/lint-and-health",1],["09-maintenance/backups-and-portability",1],["10-troubleshooting/broken-links",1]],"val":1.6},{"id":"script:vault_stats.py","type":"script","group":"tool","label":"vault_stats.py","desc":"Print health metrics for a second-brain vault.","url":"https://github.com/undefined-ui/second-brain-os/blob/main/scripts/vault_stats.py","inpages":[["04-structuring/linking-rules",1],["05-graphs/metrics",1]],"val":1.6},{"id":"plug:context-audit","type":"skill","group":"tool","label":"context-audit","sub":"plugin","desc":"Audit an agent's context layout against the four places: system prompt, tools, history, tail. Use when the user asks to audit, review or fix their agent's context, prompt caching, token spend per turn, or CLAUDE.md / AGENTS.md layout. Read-only: reports findings and a fix list, changes nothing. Do NOT use for writing…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/plugins/agents-course/skills/context-audit","inpages":[["course-1-context/context-practice",1]],"val":2.0},{"id":"plug:evals-bootstrap","type":"skill","group":"tool","label":"evals-bootstrap","sub":"plugin","desc":"Scaffold a first eval suite for an agent: mine real failures into cases, write behavioural checks over traces, and generate the runner. Use when the user wants evals, regression tests for an agent, a golden set, or asks how to know a prompt/model change did not break things. Do NOT use for a single task's done-check…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/plugins/agents-course/skills/evals-bootstrap","inpages":[["course-5-evals/evals-practice",1]],"val":2.0},{"id":"plug:gate-check","type":"skill","group":"tool","label":"gate-check","sub":"plugin","desc":"Find the decisions in a pipeline that do not need the expensive model and propose the gate for each: a rule, a classic classifier, or a small model, with fail-closed routing. Use when the user asks to cut model costs or latency, says the big model handles everything, or wants a triage / routing / filter layer in front…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/plugins/agents-course/skills/gate-check","inpages":[["course-3-gate/gate-practice",1]],"val":2.0},{"id":"plug:goal-test","type":"skill","group":"tool","label":"goal-test","sub":"plugin","desc":"Turn a vague task into a testable definition of done and generate an executable goal-test script for it, optionally with a bounded retry loop around a headless agent. Use when the user wants to run an agent in a loop, asks how to know when an agent task is finished, or says a goal like \"improve X\" needs to become…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/plugins/agents-course/skills/goal-test","inpages":[["course-2-loop/loop-practice",2]],"val":2.0},{"id":"plug:harness-audit","type":"skill","group":"tool","label":"harness-audit","sub":"plugin","desc":"Audit an agent's harness against the four rings: containment, guides, sensors, permissions. Use when the user asks whether their agent setup is safe, wants a review of sandboxing / permissions / hooks / AGENTS.md guardrails, or asks what a bad session could break. Read-only: reports the blast radius and a hardening…","url":"https://github.com/undefined-ui/second-brain-os/tree/main/plugins/agents-course/skills/harness-audit","inpages":[["course-4-harness/harness-practice",1]],"val":2.0},{"id":"plug:loop-critic","type":"agent","group":"tool","label":"loop-critic","sub":"plugin","desc":"Reviews the result of an agent loop iteration in a clean context: checks the claimed work against the goal, hunts for shortcuts and untested paths, returns a verdict with specific defects. Use after a work session or loop attempt, before accepting the result.","url":"https://github.com/undefined-ui/second-brain-os/blob/main/plugins/agents-course/agents/loop-critic.md","inpages":[["course-2-loop/loop-practice",1]],"val":2.0},{"id":"rcat:papers","type":"hub","group":"res","label":"Papers","url":"https://github.com/undefined-ui/second-brain-os/blob/main/resources/papers.md","desc":"Research relevant to how a knowledge base retrieves and reasons. Read these if you want to decide for yourself whether you need a vector database.","val":4.5},{"id":"rcat:tools","type":"hub","group":"res","label":"Tools","url":"https://github.com/undefined-ui/second-brain-os/blob/main/resources/tools.md","desc":"Everything outside the Obsidian plugin catalog. For plugins, see plugins.md. Star counts are from GitHub, September 2026.","val":4.5},{"id":"rcat:plugins","type":"hub","group":"res","label":"Obsidian plugins","url":"https://github.com/undefined-ui/second-brain-os/blob/main/resources/plugins.md","desc":"Ranked by installs from Obsidian's own community stats, September 2026. The catalog holds over 7,400 plugins; these are the ones that matter for an agent-maintained vault.","inpages":[["02-setup/obsidian-plugins",3]],"val":4.5},{"id":"rcat:repositories","type":"hub","group":"res","label":"Repositories","url":"https://github.com/undefined-ui/second-brain-os/blob/main/resources/repositories.md","desc":"Graph, retrieval and memory projects worth knowing. Stars from the GitHub API, September 2026. For agent skills see skills.md; for Obsidian plugins see plugins.md.","val":4.5},{"id":"rcat:skills","type":"hub","group":"res","label":"Skills catalogs","url":"https://github.com/undefined-ui/second-brain-os/blob/main/resources/skills.md","desc":"Counted individually: a repo shipping sixteen skills counts as sixteen, because that is what you install. Counts were taken by reading each repository's file tree in September 2026.","inpages":[["01-concepts/llm-wiki-pattern",3]],"val":4.5},{"id":"rcat:reading","type":"hub","group":"res","label":"Reading","url":"https://github.com/undefined-ui/second-brain-os/blob/main/resources/reading.md","desc":"Karpathy's llm-wiki gist, published 4 April 2026. The pattern everything in this repo builds on.","val":4.5},{"id":"res:0","type":"resource","group":"res","label":"From Local to Global: A Graph RAG Approach to Query-Focused Summarization","url":"https://arxiv.org/abs/2404.16130","metric":"","sub":"GraphRAG","desc":"Edge, Trinh, Cheng, Bradley, Chao, Mody, Truitt, Metropolitansky, Ness, Larson. Microsoft Research. Implementation at microsoft/graphrag.","inpages":[["track-graph/graphrag",1],["track-graph/resources",1]],"val":0.7},{"id":"res:1","type":"resource","group":"res","label":"HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models","url":"https://arxiv.org/abs/2405.14831","metric":"","sub":"HippoRAG","desc":"Gutiérrez, Shu, Gu, Yasunaga, Su.","inpages":[["track-graph/resources",1]],"val":0.7},{"id":"res:2","type":"resource","group":"res","label":"Lost in the Middle: How Language Models Use Long Contexts","url":"https://arxiv.org/abs/2307.03172","metric":"","sub":"Lost in the Middle","desc":"Liu, Lin, Hewitt, Paranjape, Bevilacqua, Petroni, Liang. Published in TACL.","inpages":[["course-1-context/how-models-read",1]],"val":0.7},{"id":"res:3","type":"resource","group":"res","label":"Follow the Entities: A Corpus Map for Agentic Search","url":"https://arxiv.org/abs/2609.37226","metric":"","sub":"New in October 2026","desc":"An offline pass writes one page per recurring entity that links to every document mentioning it, and agents walk that map instead of searching flat files. Across seven models and three benchmarks they found more evidence with fewer tokens.…","val":0.7},{"id":"res:4","type":"resource","group":"res","label":"Context as Infrastructure: Knowledge Compilation for Retrieval in Agentic Systems with Harnesses","url":"https://arxiv.org/abs/2610.41172","metric":"","sub":"New in October 2026","desc":"A harness with only grep and BM25 answered 85% of BrowseComp-Plus and 91% of FRAMES with no vector index. Compiled artifacts such as entity graphs and wikis added no accuracy, but a claim-evidence tree cut query tokens by 38% on average.…","val":0.7},{"id":"res:5","type":"resource","group":"res","label":"Keep Your Receipts: Trustworthy Cleanup of Agent-Built Knowledge Graphs","url":"https://arxiv.org/abs/2610.17719","metric":"","sub":"New in October 2026","desc":"Every merge, edge and summary sentence the model writes records the evidence behind it, so deleting a retracted source becomes set arithmetic with no model calls. Two builds of the same 300 documents differed in 29,475 artifacts, which is…","val":0.7},{"id":"res:6","type":"resource","group":"res","label":"LiteRAG: Cost-Efficient Graph-Based Retrieval-Augmented Generation","url":"https://arxiv.org/abs/2609.10239","metric":"","sub":"New in October 2026","desc":"Graph retrieval steered by algorithms at query time, with adaptive thresholds and a penalty on hub nodes, instead of an LLM on every hop. Against GraphRAG Global and DRIFT it cut latency over 100x and cost over 99% while matching quality.","val":0.7},{"id":"res:7","type":"resource","group":"res","label":"EvalMem: An Operation-Level Diagnostic Framework for Long-Term Memory Systems","url":"https://arxiv.org/abs/2609.22231","metric":"","sub":"New in October 2026","desc":"Splits every memory failure into not stored, not retrieved or not used, and retrieval was the most common. Adding a wiki-style layer of source pages, hub pages and versioned facts improved accuracy by 2.5 points on LoCoMo.","val":0.7},{"id":"res:8","type":"resource","group":"res","label":"CliffCompaction: Cost-Efficient Compaction for Long-Horizon Coding Agents","url":"https://arxiv.org/abs/2609.26779","metric":"","sub":"New in October 2026","desc":"Compaction that only truncates or drops content and never compacts an earlier compaction. Up to 50% cheaper under a bounded context, with Terminal-Bench scores held or improved, and simple enough to copy.","val":0.7},{"id":"res:9","type":"resource","group":"res","label":"The Unreliable Progress Bar: Can LLM Agents Reliably Report Task Progress Throughout Execution?","url":"https://arxiv.org/abs/2609.08589","metric":"","sub":"New in October 2026","desc":"Most models lose accuracy reporting their own progress mid-task, and the newest turn conservative near the finish. The authors conclude that frameworks should not drive task flow on self-reports alone, which is the case for stop rules…","val":0.7},{"id":"res:10","type":"resource","group":"res","label":"Frozen Judges, Moving Agents: Version-Dependent LLM-Judge Error","url":"https://arxiv.org/abs/2609.34198","metric":"","sub":"New in October 2026","desc":"A fixed LLM judge makes different errors on different agent versions. On SWE-bench Verified, eight judge-only comparisons declared upgrades that execution results could not confirm. Do not sign off an upgrade on judge scores alone.","val":0.7},{"id":"res:11","type":"resource","group":"res","label":"Obsidian","url":"https://obsidian.md","metric":"","sub":"The core stack","desc":"Local markdown editor. The vault is a plain folder, which is why agents work well with it","inpages":[["02-setup/obsidian-install-and-vault",1]],"val":0.7},{"id":"res:12","type":"resource","group":"res","label":"Obsidian Web Clipper","url":"https://obsidian.md/clipper","metric":"","sub":"The core stack","desc":"Official browser extension, saves articles into `raw/`. Uses Mozilla Readability","inpages":[["03-ingestion/web-clipper",1]],"val":0.7},{"id":"res:13","type":"resource","group":"res","label":"Claude Code","url":"https://code.claude.com/docs/en/setup","metric":"","sub":"The core stack","desc":"The agent that maintains the wiki. Paid account required","inpages":[["02-setup/claude-code-setup",1]],"val":0.7},{"id":"res:14","type":"resource","group":"res","label":"AFFiNE","url":"https://github.com/toeverything/AFFiNE","metric":"72,292","sub":"Alternative homes for a vault","desc":"Docs, whiteboard and database in one. Heavier than markdown files","val":0.7},{"id":"res:15","type":"resource","group":"res","label":"memos","url":"https://github.com/usememos/memos","metric":"62,837","sub":"Alternative homes for a vault","desc":"Self-hosted quick capture, markdown-native. A good `raw/` inbox","val":0.7},{"id":"res:16","type":"resource","group":"res","label":"Joplin","url":"https://github.com/laurent22/joplin","metric":"56,264","sub":"Alternative homes for a vault","desc":"Privacy-focused notes with sync across every platform","val":0.7},{"id":"res:17","type":"resource","group":"res","label":"SiYuan","url":"https://github.com/siyuan-note/siyuan","metric":"46,225","sub":"Alternative homes for a vault","desc":"Self-hosted knowledge workspace, block-based","val":0.7},{"id":"res:18","type":"resource","group":"res","label":"Logseq","url":"https://github.com/logseq/logseq","metric":"44,818","sub":"Alternative homes for a vault","desc":"Outliner, plain markdown, block-level references","val":0.7},{"id":"res:19","type":"resource","group":"res","label":"Outline","url":"https://github.com/outline/outline","metric":"40,483","sub":"Alternative homes for a vault","desc":"Team knowledge base. The right answer when it is not just you","val":0.7},{"id":"res:20","type":"resource","group":"res","label":"Trilium","url":"https://github.com/TriliumNext/Trilium","metric":"37,762","sub":"Alternative homes for a vault","desc":"Hierarchical personal knowledge base with scripting","val":0.7},{"id":"res:21","type":"resource","group":"res","label":"Foam","url":"https://github.com/foambubble/foam","metric":"17,386","sub":"Alternative homes for a vault","desc":"The same wikilink pattern inside VS Code","val":0.7},{"id":"res:22","type":"resource","group":"res","label":"Anytype","url":"https://github.com/anyproto/anytype-ts","metric":"8,764","sub":"Alternative homes for a vault","desc":"Local-first, encrypted, object-based","val":0.7},{"id":"res:23","type":"resource","group":"res","label":"Dendron","url":"https://github.com/dendronhq/dendron","metric":"7,465","sub":"Alternative homes for a vault","desc":"Hierarchical PKM in VS Code","val":0.7},{"id":"res:24","type":"resource","group":"res","label":"Khoj","url":"https://github.com/khoj-ai/khoj","metric":"37,186","sub":"AI-native knowledge apps","desc":"Self-hostable AI second brain over your docs, with custom agents and scheduled automations","val":0.7},{"id":"res:25","type":"resource","group":"res","label":"Quivr","url":"https://github.com/The-Vibe-Company/quivr","metric":"39,498","sub":"AI-native knowledge apps","desc":"Opinionated RAG you embed in your own app","val":0.7},{"id":"res:26","type":"resource","group":"res","label":"Reor","url":"https://github.com/reorproject/reor","metric":"8,563","sub":"AI-native knowledge apps","desc":"Local-first AI note app that links notes automatically as you write","val":0.7},{"id":"res:27","type":"resource","group":"res","label":"yt-dlp","url":"https://github.com/yt-dlp/yt-dlp","metric":"","sub":"Capture and processing","desc":"Subtitles and audio from video platforms","val":0.7},{"id":"res:28","type":"resource","group":"res","label":"youtube-transcript-api","url":"https://github.com/jdepoix/youtube-transcript-api","metric":"","sub":"Capture and processing","desc":"Transcripts from Python","val":0.7},{"id":"res:29","type":"resource","group":"res","label":"OCRmyPDF","url":"https://github.com/ocrmypdf/OCRmyPDF","metric":"","sub":"Capture and processing","desc":"Text layer for scanned PDFs","val":0.7},{"id":"res:30","type":"resource","group":"res","label":"Zotero","url":"https://github.com/zotero/zotero","metric":"","sub":"Capture and processing","desc":"Reference manager, 15,184 stars, markdown export path","val":0.7},{"id":"res:31","type":"resource","group":"res","label":"Readwise","url":"https://readwise.io","metric":"","sub":"Capture and processing","desc":"Kindle and read-later highlights into Obsidian. Paid","val":0.7},{"id":"res:32","type":"resource","group":"res","label":"ripgrep","url":"https://github.com/BurntSushi/ripgrep","metric":"","sub":"Search and graph","desc":"Fast enough that most vaults never need an index","val":0.7},{"id":"res:33","type":"resource","group":"res","label":"NetworkX","url":"https://networkx.org","metric":"","sub":"Search and graph","desc":"Graph analysis once you have exported an edge list","inpages":[["track-graph/graph-stores",1]],"val":0.7},{"id":"res:34","type":"resource","group":"res","label":"Ladybug","url":"https://github.com/LadybugDB/ladybug","metric":"","sub":"Search and graph","desc":"Embedded graph database, no server. The maintained continuation of Kuzu, which was archived in October 2025","val":0.7},{"id":"res:35","type":"resource","group":"res","label":"Neo4j","url":"https://neo4j.com","metric":"","sub":"Search and graph","desc":"Full graph database. Overkill unless you build on top","inpages":[["track-graph/graph-stores",1]],"val":0.7},{"id":"res:36","type":"resource","group":"res","label":"Gephi","url":"https://gephi.org","metric":"","sub":"Search and graph","desc":"Visual graph exploration beyond Obsidian's view","inpages":[["track-graph/tools",1]],"val":0.7},{"id":"res:37","type":"resource","group":"res","label":"txtai","url":"https://github.com/neuml/txtai","metric":"","sub":"Search and graph","desc":"12,931 stars. Semantic search and LLM workflows, if you decide you need embeddings","val":0.7},{"id":"res:38","type":"resource","group":"res","label":"Quartz","url":"https://github.com/jackyzha0/quartz","metric":"13,185","sub":"Publishing","desc":"Vault to static site with backlinks, local graph and search","val":0.7},{"id":"res:39","type":"resource","group":"res","label":"Claudian","url":"https://github.com/yishentu/claudian","metric":"2,035,966","sub":"Agents inside Obsidian","desc":"Embeds Claude Code, Codex and other local agents as collaborators in the vault","inpages":[["02-setup/claude-code-setup",1]],"val":0.7},{"id":"res:40","type":"resource","group":"res","label":"Copilot","url":"https://github.com/logancyang/obsidian-copilot","metric":"1,832,464","sub":"Agents inside Obsidian","desc":"Runs Claude Code, Codex and OpenCode inside your vault","inpages":[["02-setup/claude-code-setup",1]],"val":0.7},{"id":"res:41","type":"resource","group":"res","label":"Smart Connections","url":"https://github.com/brianpetro/obsidian-smart-connections","metric":"1,189,513","sub":"Agents inside Obsidian","desc":"Local embedding model surfaces related notes while you write. No API key","val":0.7},{"id":"res:42","type":"resource","group":"res","label":"Local REST API with MCP","url":"https://github.com/coddingtonbear/obsidian-local-rest-api","metric":"712,443","sub":"Agents inside Obsidian","desc":"Vault over a secure local API, now with a built-in MCP server","inpages":[["02-setup/mcp-obsidian",1]],"val":0.7},{"id":"res:43","type":"resource","group":"res","label":"Agent Client","url":"https://github.com/rait-09/obsidian-agent-client","metric":"261,371","sub":"Agents inside Obsidian","desc":"Chat with Claude Code, Codex and Gemini CLI over the Agent Client Protocol","val":0.7},{"id":"res:44","type":"resource","group":"res","label":"Smart Composer","url":"https://github.com/glowingjade/obsidian-smart-composer","metric":"171,863","sub":"Agents inside Obsidian","desc":"AI chat with note context and one-click edits","val":0.7},{"id":"res:45","type":"resource","group":"res","label":"Templater","url":"https://github.com/SilentVoid13/Templater","metric":"5,527,936","sub":"Structure and queries","desc":"Dynamic templates. For the pages you write by hand","val":0.7},{"id":"res:46","type":"resource","group":"res","label":"Dataview","url":"https://github.com/blacksmithgu/obsidian-dataview","metric":"4,918,446","sub":"Structure and queries","desc":"Query language over your frontmatter. Answers \"which pages have property P\"","val":0.7},{"id":"res:47","type":"resource","group":"res","label":"Metadata Menu","url":"https://github.com/mdelobelle/metadatamenu","metric":"330,430","sub":"Structure and queries","desc":"Manage frontmatter fields at scale. Useful once your schema is fixed","val":0.7},{"id":"res:48","type":"resource","group":"res","label":"Supercharged Links","url":"https://github.com/mdelobelle/obsidian_supercharged_links","metric":"210,926","sub":"Structure and queries","desc":"Styles links by the target note's frontmatter, so page type is visible inline","val":0.7},{"id":"res:49","type":"resource","group":"res","label":"ExcaliBrain","url":"https://github.com/zsviczian/excalibrain","metric":"336,025","sub":"Graph","desc":"Structured mind-map deriving five relationship types from links, Dataview fields and tags","val":0.7},{"id":"res:50","type":"resource","group":"res","label":"Juggl","url":"https://github.com/HEmile/juggl","metric":"135,614","sub":"Graph","desc":"Interactive Cytoscape.js graph with typed edges and a saveable workspace mode","val":0.7},{"id":"res:51","type":"resource","group":"res","label":"Graph Analysis","url":"https://github.com/SkepticMystic/graph-analysis","metric":"community plugin","sub":"Graph","desc":"Graph algorithms and similarity measures over the vault, surfaces unlinked connections","val":0.7},{"id":"res:52","type":"resource","group":"res","label":"Breadcrumbs","url":"https://github.com/michaelpporter/breadcrumbs","metric":"community plugin","sub":"Graph","desc":"Typed links plus trees, matrices, Mermaid and Canvas export","val":0.7},{"id":"res:53","type":"resource","group":"res","label":"3D Graph","url":"https://github.com/AlexW00/obsidian-3d-graph","metric":"community plugin","sub":"Graph","desc":"The vault as a 3D force graph. Shows cluster structure a 2D hairball hides","val":0.7},{"id":"res:54","type":"resource","group":"res","label":"Git","url":"https://github.com/Vinzent03/obsidian-git","metric":"3,110,571","sub":"Maintenance","desc":"Commit and sync the vault from inside Obsidian. See versioning","inpages":[["02-setup/obsidian-plugins",1]],"val":0.7},{"id":"res:55","type":"resource","group":"res","label":"Omnisearch","url":"https://github.com/scambier/obsidian-omnisearch","metric":"1,837,484","sub":"Maintenance","desc":"Better full-text search, including inside PDFs","inpages":[["02-setup/obsidian-plugins",1]],"val":0.7},{"id":"res:56","type":"resource","group":"res","label":"Importer","url":"https://github.com/obsidianmd/obsidian-importer","metric":"1,638,001","sub":"Maintenance","desc":"Official importer from Notion, Evernote, Roam, Bear, Apple Notes. The first step for anyone backfilling","inpages":[["02-setup/obsidian-plugins",1],["10-troubleshooting/faq",1]],"val":0.7},{"id":"res:57","type":"resource","group":"res","label":"Linter","url":"https://github.com/platers/obsidian-linter","metric":"1,044,521","sub":"Maintenance","desc":"Formats notes to consistent rules. Set it carefully, it rewrites files","val":0.7},{"id":"res:58","type":"resource","group":"res","label":"Find orphaned files and broken links","url":"https://github.com/Vinzent03/find-unlinked-files","metric":"224,687","sub":"Maintenance","desc":"Exactly what lint checks, without leaving Obsidian","inpages":[["02-setup/obsidian-plugins",1],["09-maintenance/lint-and-health",1]],"val":0.7},{"id":"res:59","type":"resource","group":"res","label":"Excalidraw","url":"https://github.com/zsviczian/obsidian-excalidraw-plugin","metric":"7,795,993","sub":"Capture and study","desc":"The most installed plugin in the catalog. Diagrams that live in the vault as files","val":0.7},{"id":"res:60","type":"resource","group":"res","label":"QuickAdd","url":"https://github.com/chhoumann/quickadd","metric":"2,082,182","sub":"Capture and study","desc":"One-keystroke capture into a chosen folder and template","val":0.7},{"id":"res:61","type":"resource","group":"res","label":"Spaced Repetition","url":"https://github.com/st3v3nmw/obsidian-spaced-repetition","metric":"592,497","sub":"Capture and study","desc":"Flashcards from your own notes. See learning from your vault","inpages":[["08-outputs/teaching-yourself",1]],"val":0.7},{"id":"res:62","type":"resource","group":"res","label":"Zotero Integration","url":"https://github.com/obsidian-community/obsidian-zotero-integration","metric":"550,881","sub":"Capture and study","desc":"Citations, bibliographies and PDF annotations from Zotero","val":0.7},{"id":"res:63","type":"resource","group":"res","label":"ReadItLater","url":"https://github.com/DominikPieper/obsidian-ReadItLater","metric":"134,982","sub":"Capture and study","desc":"Saves online content into the vault with a template","val":0.7},{"id":"res:64","type":"resource","group":"res","label":"Webpage HTML Export","url":"https://github.com/KosmosisDire/obsidian-webpage-export","metric":"152,194","sub":"Capture and study","desc":"Export notes, canvases or the whole vault as HTML","val":0.7},{"id":"res:65","type":"resource","group":"res","label":"HKUDS/LightRAG","url":"https://github.com/HKUDS/LightRAG","metric":"39,467","sub":"GraphRAG and graph retrieval","desc":"Dual-layer graph plus vectors, aimed at cheap incremental updates. EMNLP 2025. Since 1.5.7 (September 2026) it can build the graph first and add vectors later","inpages":[["track-graph/why-graphs",1],["track-graph/graphrag",1],["track-graph/tools",1],["track-graph/resources",1]],"val":0.7},{"id":"res:66","type":"resource","group":"res","label":"microsoft/graphrag","url":"https://github.com/microsoft/graphrag","metric":"35,875","sub":"GraphRAG and graph retrieval","desc":"The reference implementation of the paper","inpages":[["track-graph/graphrag",1],["track-graph/tools",1],["track-graph/resources",1]],"val":0.7},{"id":"res:67","type":"resource","group":"res","label":"gusye1234/nano-graphrag","url":"https://github.com/gusye1234/nano-graphrag","metric":"3,983","sub":"GraphRAG and graph retrieval","desc":"The same idea in about 1,100 readable lines. Open this one to understand the pipeline","val":0.7},{"id":"res:68","type":"resource","group":"res","label":"OSU-NLP-Group/HippoRAG","url":"https://github.com/OSU-NLP-Group/HippoRAG","metric":"3,985","sub":"GraphRAG and graph retrieval","desc":"The NeurIPS'24 paper's implementation. Personalised PageRank over a knowledge graph","val":0.7},{"id":"res:69","type":"resource","group":"res","label":"DEEP-PolyU/Awesome-GraphRAG","url":"https://github.com/DEEP-PolyU/Awesome-GraphRAG","metric":"2,635","sub":"GraphRAG and graph retrieval","desc":"Curated map of surveys, benchmarks and projects in the area","val":0.7},{"id":"res:70","type":"resource","group":"res","label":"mem0ai/mem0","url":"https://github.com/mem0ai/mem0","metric":"64,863","sub":"Agent memory","desc":"Drop-in memory layer for agents and apps","val":0.7},{"id":"res:71","type":"resource","group":"res","label":"getzep/graphiti","url":"https://github.com/getzep/graphiti","metric":"30,680","sub":"Agent memory","desc":"Real-time temporal knowledge graphs for agents","inpages":[["track-graph/why-graphs",1],["track-graph/building-graphs-with-llms",1],["track-graph/graph-stores",1],["track-graph/tools",1],["track-graph/resources",1]],"val":0.7},{"id":"res:72","type":"resource","group":"res","label":"topoteretes/cognee","url":"https://github.com/topoteretes/cognee","metric":"30,568","sub":"Agent memory","desc":"Turns data into knowledge graphs, combining graph and vector retrieval","inpages":[["track-graph/tools",1]],"val":0.7},{"id":"res:73","type":"resource","group":"res","label":"doobidoo/mcp-memory-service","url":"https://github.com/doobidoo/mcp-memory-service","metric":"1,928","sub":"Agent memory","desc":"Persistent memory over MCP, REST API plus knowledge graph","val":0.7},{"id":"res:74","type":"resource","group":"res","label":"infiniflow/ragflow","url":"https://github.com/infiniflow/ragflow","metric":"90,216","sub":"RAG frameworks","desc":"Full RAG engine with document understanding","val":0.7},{"id":"res:75","type":"resource","group":"res","label":"run-llama/llama_index","url":"https://github.com/run-llama/llama_index","metric":"52,058","sub":"RAG frameworks","desc":"Document agent and ingestion framework","val":0.7},{"id":"res:76","type":"resource","group":"res","label":"neuml/txtai","url":"https://github.com/neuml/txtai","metric":"12,931","sub":"RAG frameworks","desc":"Embeddings database and LLM workflows, small enough to read","val":0.7},{"id":"res:77","type":"resource","group":"res","label":"punkpeye/awesome-mcp-servers","url":"https://github.com/punkpeye/awesome-mcp-servers","metric":"94,575","sub":"MCP","desc":"The index of MCP servers","val":0.7},{"id":"res:78","type":"resource","group":"res","label":"modelcontextprotocol/servers","url":"https://github.com/modelcontextprotocol/servers","metric":"90,141","sub":"MCP","desc":"The official reference servers, including filesystem and memory","val":0.7},{"id":"res:79","type":"resource","group":"res","label":"MarkusPfundstein/mcp-obsidian","url":"https://github.com/MarkusPfundstein/mcp-obsidian","metric":"4,376","sub":"MCP","desc":"Third-party MCP server for Obsidian over the REST plugin","val":0.7},{"id":"res:80","type":"resource","group":"res","label":"basicmachines-co/basic-memory","url":"https://github.com/basicmachines-co/basic-memory","metric":"4,445","sub":"New in October 2026","desc":"A local MCP server whose knowledge graph is plain markdown with typed `relation [[WikiLink]]` lines, indexed into SQLite. Reads and writes Obsidian vaults. 1.0 shipped on 29 October. AGPL-3.0","val":0.7},{"id":"res:81","type":"resource","group":"res","label":"akitaonrails/ai-memory","url":"https://github.com/akitaonrails/ai-memory","metric":"14,856","sub":"New in October 2026","desc":"A Rust memory server whose source of truth is a git-backed wiki of ordinary `.md` files, fed by hooks from 20+ agent harnesses. 3.0 (13 October) installs as a Claude Code plugin marketplace. MIT","val":0.7},{"id":"res:82","type":"resource","group":"res","label":"tigerless-labs/agent-memory","url":"https://github.com/tigerless-labs/agent-memory","metric":"4,306","sub":"New in October 2026","desc":"One markdown file per memory under a `MEMORY.md` index. The SQLite and BM25 index is a cache you can delete and rebuild, and Claude Code and Codex share one store. MIT","val":0.7},{"id":"res:83","type":"resource","group":"res","label":"vectorize-io/hindsight","url":"https://github.com/vectorize-io/hindsight","metric":"57,739","sub":"New in October 2026","desc":"The database-backed alternative: retain, recall and reflect over vector, BM25, graph and time search, with knowledge pages it can write out as markdown. 1.0 on 13 October. MIT","val":0.7},{"id":"res:84","type":"resource","group":"res","label":"decodingai-magazine/llm-wiki-workshop","url":"https://github.com/decodingai-magazine/llm-wiki-workshop","metric":"43","sub":"New in October 2026","desc":"A self-paced workshop: three exercises, each a working system, from a basic `raw/` to `wiki/` setup to a wiki that learns from queries. MIT","val":0.7},{"id":"res:85","type":"resource","group":"res","label":"LadybugDB/ladybug","url":"https://github.com/LadybugDB/ladybug","metric":"1,895","sub":"New in October 2026","desc":"An embedded, serverless Cypher graph database, formerly Kuzu, which was archived in October 2025. 0.22 on 25 October. MIT","inpages":[["05-graphs/exporting-your-graph",1]],"val":0.7},{"id":"res:86","type":"resource","group":"res","label":"google-research/harness2","url":"https://github.com/google-research/harness2","metric":"21,920","sub":"New in October 2026","desc":"Harness², recursive harness tuning: the weights stay frozen while prompts, tools and memory are edited per task and the lessons kept. Apache-2.0","val":0.7},{"id":"res:87","type":"resource","group":"res","label":"NVlabs/SoL-Pi","url":"https://github.com/NVlabs/SoL-Pi","metric":"3,620","sub":"New in October 2026","desc":"Four opt-in add-ons for the Pi harness that keep its score with 45 to 49 per cent less token traffic. MIT","val":0.7},{"id":"res:88","type":"resource","group":"res","label":"AgriciDaniel/claude-obsidian","url":"https://github.com/AgriciDaniel/claude-obsidian","metric":"14,706","sub":"Implementations of this pattern","desc":"16 skills, 3 subagents. Self-organizing vault with role presets","val":0.7},{"id":"res:89","type":"resource","group":"res","label":"Astro-Han/karpathy-llm-wiki","url":"https://github.com/Astro-Han/karpathy-llm-wiki","metric":"2,174","sub":"Implementations of this pattern","desc":"1 skill covering the full ingest, compile, query, lint loop","val":0.7},{"id":"res:90","type":"resource","group":"res","label":"ballred/obsidian-claude-pkm","url":"https://github.com/ballred/obsidian-claude-pkm","metric":"1,856","sub":"Implementations of this pattern","desc":"13 skills, 4 subagents. A complete starter kit","val":0.7},{"id":"res:91","type":"resource","group":"res","label":"coleam00/second-brain-starter","url":"https://github.com/coleam00/second-brain-starter","metric":"768","sub":"Implementations of this pattern","desc":"1 skill that interviews you and generates a build plan","val":0.7},{"id":"res:92","type":"resource","group":"res","label":"NicholasSpisak/second-brain","url":"https://github.com/NicholasSpisak/second-brain","metric":"704","sub":"Implementations of this pattern","desc":"4 skills, npm installer, close to the original gist","val":0.7},{"id":"res:93","type":"resource","group":"res","label":"eugeniughelbur/obsidian-second-brain","url":"https://github.com/eugeniughelbur/obsidian-second-brain","metric":"small","sub":"Implementations of this pattern","desc":"47 commands across Claude, Codex and Gemini","val":0.7},{"id":"res:94","type":"resource","group":"res","label":"micuintus/llm-wiki","url":"https://github.com/micuintus/llm-wiki","metric":"small","sub":"Implementations of this pattern","desc":"1 skill, deliberately minimal, no dependencies. Good counterpoint","val":0.7},{"id":"res:95","type":"resource","group":"res","label":"obra/superpowers","url":"https://github.com/obra/superpowers","metric":"282,827","sub":"General skill libraries","desc":"14 skills. An agentic skills framework and development methodology","val":0.7},{"id":"res:96","type":"resource","group":"res","label":"anthropics/skills","url":"https://github.com/anthropics/skills","metric":"175,035","sub":"General skill libraries","desc":"20 skills, 3 subagents. The official reference for the format","val":0.7},{"id":"res:97","type":"resource","group":"res","label":"VoltAgent/awesome-agent-skills","url":"https://github.com/VoltAgent/awesome-agent-skills","metric":"33,894","sub":"General skill libraries","desc":"A curated index of 1,000+ community skills","val":0.7},{"id":"res:98","type":"resource","group":"res","label":"hesreallyhim/awesome-claude-code","url":"https://github.com/hesreallyhim/awesome-claude-code","metric":"53,661","sub":"General skill libraries","desc":"Commands, hooks, workflows and tooling for Claude Code","val":0.7},{"id":"res:99","type":"resource","group":"res","label":"Claude Code mods","url":"https://code.claude.com/docs/en/plugins/mods/overview","metric":"","sub":"New in October 2026","desc":"A new plugin part beside skills, hooks and MCP: JS or TS handlers that watch or rewrite tool calls and prompts, add commands and draw panes. Not sandboxed, so read one before you install it","val":0.7},{"id":"res:100","type":"resource","group":"res","label":"claude plugin eval","url":"https://code.claude.com/docs/en/plugin-evals","metric":"","sub":"New in October 2026","desc":"Runs each eval case with and without your plugin, scores the difference and exits non-zero below a threshold, so CI can gate on it","inpages":[["track-evals/tooling",1]],"val":0.7},{"id":"res:101","type":"resource","group":"res","label":"/skill-doctor","url":"https://code.claude.com/docs/en/skills#find-unused-skills","metric":"","sub":"New in October 2026","desc":"Reports what each loaded skill costs in context on every turn and which ones are never called","val":0.7},{"id":"res:102","type":"resource","group":"res","label":"AGENTS.md support","url":"https://code.claude.com/docs/en/memory#agents-md","metric":"","sub":"New in October 2026","desc":"Claude Code reads `AGENTS.md` when there is no `CLAUDE.md`, and a setting loads both, so one instruction file can serve several agents","val":0.7},{"id":"res:103","type":"resource","group":"res","label":"google/elephant-goldfish-model","url":"https://github.com/google/elephant-goldfish-model","metric":"9,069","sub":"New in October 2026","desc":"A plugin of seven skills: a full-context agent drafts a spec and reviewers with no prior conversation check it. Installs in Claude Code, Codex and Antigravity. Apache-2.0","val":0.7},{"id":"res:104","type":"resource","group":"res","label":"Graphify-Labs/graphify","url":"https://github.com/Graphify-Labs/graphify","metric":"115,697","sub":"Graph builders","desc":"A `/graphify` skill that turns any folder of code, docs, PDFs and screenshots into a queryable graph. Every edge labelled extracted or inferred","val":0.7},{"id":"res:105","type":"resource","group":"res","label":"Karpathy's llm-wiki gist","url":"https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f","metric":"","sub":"The source document","desc":"The pattern everything in this repo builds on.","val":0.7},{"id":"res:106","type":"resource","group":"res","label":"An organizational second brain: AI that learns from experts","url":"https://engineering.fb.com/2026/09/02/ml-applications/organizational-second-brain-ai-learns-from-experts/","metric":"","sub":"The pattern at company scale","desc":"The same pattern this guide teaches, run inside Meta with a whole domain team as the user.","val":0.7},{"id":"res:107","type":"resource","group":"res","label":"Andy Matuschak's notes","url":"https://notes.andymatuschak.org","metric":"","sub":"The tradition it comes from","desc":"The Zettelkasten method. The argument for atomic notes and dense linking still holds; the manual labour it demands is the part that killed it for most people. - Building a Second Brain, Tiago Forte. Where the term comes from, and the PARA…","val":0.7},{"id":"res:108","type":"resource","group":"res","label":"Anthropic donates Agent Skills to the Agentic AI Foundation","url":"https://www.anthropic.com/news/donating-agent-skills-to-the-agentic-ai-foundation","metric":"","sub":"New in October 2026","desc":"The `SKILL.md` format moves to the Linux Foundation's agent foundation, name, spec and site included. The skills in this repo are now written to a neutral standard.","val":0.7},{"id":"res:109","type":"resource","group":"res","label":"Building agents: when to fold subagents into your main agent loop","url":"https://claude.com/resources/articles/building-agents-when-to-fold-subagents-into-your-main-agent-loop","metric":"","sub":"New in October 2026","desc":"The case that one loop with skills and progressively disclosed tools often matches a fleet of subagents, when subagents are still worth it, and a migration plan that starts from ten golden eval cases.","val":0.7},{"id":"res:110","type":"resource","group":"res","label":"Building effective agent automations","url":"https://claude.dev/blog/building-effective-agent-automations/","metric":"","sub":"New in October 2026","desc":"A scheduled agent built end to end: a bookmark per source, a ledger of what was already reported, a per-run spend cap, and a failed source reported as a failure rather than a quiet day. Every pattern carries over to a nightly ingest.","val":0.7},{"id":"res:111","type":"resource","group":"res","label":"Automating eval design and hillclimbing with Claude","url":"https://claude.dev/blog/automating-eval-design-and-hillclimbing/","metric":"","sub":"New in October 2026","desc":"Two commands that build an eval in your codebase and then improve against it, keeping a change only when both the train split and a held-out split improve. Read it with Hamel Husain's critique.","val":0.7},{"id":"res:112","type":"resource","group":"res","label":"How to use decision models for evals","url":"https://hamel.dev/blog/posts/decision-model-evals/","metric":"","sub":"New in October 2026","desc":"Jev, OpenAI's Decisions API and three other decision models compared as cheap pass or fail judges on 98 human-labelled agent traces, with the advice to tune the threshold rather than keep 0.5.","val":0.7},{"id":"res:113","type":"resource","group":"res","label":"How the Claude Code team does code review","url":"https://claude.dev/blog/solving-code-review-at-anthropic/","metric":"","sub":"New in October 2026","desc":"Parallel finders, a separate voting verifier and a three-round cap took first-round passes from 65% to 78%. Real numbers on making a critic loop converge.","val":0.7},{"id":"term:second-brain","type":"term","group":"term","label":"Second brain","desc":"One local folder of plain text that an agent reads and writes, holding what you have read and understood, that you can ask questions of.","inpages":[["01-concepts/what-is-a-second-brain",0],["track-graph/why-graphs",1],["track-harness/context-engineering",1],["track-loop/context-hygiene",1]],"total":3,"val":1.7},{"id":"term:llm-wiki","type":"term","group":"term","label":"LLM wiki","desc":"Karpathy's pattern: the model compiles sources into a linked wiki once, instead of rediscovering them from chunks on every question.","inpages":[["01-concepts/llm-wiki-pattern",0],["01-concepts/the-save-for-later-paradox",1],["01-concepts/two-layers",1]],"total":2,"val":1.62},{"id":"term:vault","type":"term","group":"term","label":"Vault","desc":"The folder of markdown the agent maintains. Obsidian calls it a vault; to the agent it is just files.","inpages":[["07-retrieval/asking-questions",9],["02-setup/claude-code-setup",8],["10-troubleshooting/common-failures",7],["09-maintenance/privacy-and-secrets",6],["02-setup/obsidian-install-and-vault",5],["02-setup/vault-structure",4],["08-outputs/writing-from-the-vault",4],["05-graphs/graphrag",3],["06-agents/safety-and-guardrails",3],["07-retrieval/context-budget",3],["05-graphs/graph-vs-vectors",2],["08-outputs/teaching-yourself",2],["09-maintenance/lint-and-health",2],["10-troubleshooting/vault-too-big",2],["07-retrieval/query-patterns",10],["02-setup/claude-md",9],["02-setup/project-scoping",7],["05-graphs/metrics",7],["08-outputs/publishing-and-export",7],["02-setup/mcp-obsidian",6],["05-graphs/graph-basics",6],["08-outputs/research-reports",6],["track-graph/build-extract",5],["track-graph/build-use",5],["02-setup/git-and-sync",5],["06-agents/agent-roles",5],["06-agents/skills-and-commands",5],["10-troubleshooting/faq",5],["02-setup/obsidian-plugins",4],["03-ingestion/voice-and-meetings",4],["04-structuring/page-types",4],["04-structuring/frontmatter-schema",4],["05-graphs/exporting-your-graph",4],["07-retrieval/rag-on-top",4],["09-maintenance/scaling",4],["track-graph/build-query",3],["01-concepts/why-markdown-and-plain-text",3],["01-concepts/what-good-looks-like",3],["02-setup/live-data",3],["03-ingestion/chat-exports",3],["03-ingestion/newsletters-and-email",3],["04-structuring/linking-rules",3],["04-structuring/naming-and-aliases",3],["04-structuring/dedupe-and-merge",3],["04-structuring/contradictions-and-supersession",3],["04-structuring/index-and-log",3],["09-maintenance/backups-and-portability",3],["10-troubleshooting/broken-links",3],["track-graph/tools",2],["03-ingestion/bulk-backfill",2],["06-agents/scheduled-maintenance",2],["09-maintenance/review-cadence",2],["09-maintenance/versioning-with-git",2]],"total":225,"val":3.28},{"id":"term:obsidian","type":"term","group":"term","label":"Obsidian","desc":"The markdown editor used to read the vault, browse its graph and clip sources. The agent never needs it running.","inpages":[["02-setup/mcp-obsidian",7],["02-setup/obsidian-install-and-vault",3],["02-setup/claude-code-setup",3],["track-graph/tools",2],["03-ingestion/web-clipper",2],["07-retrieval/search-tools",1],["05-graphs/obsidian-graph-view",0],["02-setup/git-and-sync",5],["02-setup/obsidian-plugins",3],["10-troubleshooting/faq",3],["05-graphs/exporting-your-graph",2],["08-outputs/publishing-and-export",2],["09-maintenance/versioning-with-git",2],["09-maintenance/backups-and-portability",2]],"total":37,"val":2.22},{"id":"term:claude-code","type":"term","group":"term","label":"Claude Code","desc":"The agent that does the filing: reads raw sources, writes and links pages, runs the maintenance.","inpages":[["02-setup/claude-code-setup",6],["track-harness/claude-code-as-harness",2],["track-evals/tooling",2],["track-harness/build-graduate",3],["course-4-harness/the-four-rings",2],["track-graph/build-extract",2],["track-harness/harness-landscape",2],["10-troubleshooting/faq",2]],"total":21,"val":1.95},{"id":"term:claude-md","type":"term","group":"term","label":"CLAUDE.md","desc":"Standing instructions the agent reads at the start of every session: who you are, the page contracts, the vault's rules.","inpages":[["02-setup/claude-md",7],["track-harness/claude-code-as-harness",2],["10-troubleshooting/agent-writes-garbage",2],["02-setup/vault-structure",4],["02-setup/claude-code-setup",3],["02-setup/project-scoping",3],["10-troubleshooting/common-failures",2]],"total":23,"val":1.89},{"id":"term:mcp","type":"term","group":"term","label":"MCP","desc":"Model Context Protocol: the standard plug that lets an agent reach tools and data, such as your vault from anywhere.","inpages":[["02-setup/claude-code-setup",5],["track-harness/claude-code-as-harness",4],["track-harness/tools-and-mcp",4],["02-setup/mcp-obsidian",3],["course-4-harness/the-four-rings",3],["track-harness/resources",3],["10-troubleshooting/faq",3],["02-setup/live-data",2],["04-structuring/naming-and-aliases",2]],"total":29,"val":2.0},{"id":"term:raw","type":"term","group":"term","label":"raw/","desc":"The inbox and archive: clipped articles, transcripts, PDFs. Written once, never edited, rarely read again.","inpages":[["06-agents/scheduled-maintenance",3],["02-setup/vault-structure",2],["03-ingestion/web-clipper",2],["03-ingestion/chat-exports",2],["03-ingestion/newsletters-and-email",2]],"total":11,"val":1.77},{"id":"term:two-layers","type":"term","group":"term","label":"Two layers","desc":"A wiki for what you know and a project layer for what you are doing, kept apart so each stays useful.","inpages":[["01-concepts/two-layers",5],["02-setup/vault-structure",5],["01-concepts/llm-wiki-pattern",3],["03-ingestion/voice-and-meetings",2]],"total":15,"val":1.7},{"id":"term:ingest","type":"term","group":"term","label":"Ingest","desc":"Turning one raw source into linked wiki pages. Not finished until the new pages are connected to what was already there.","inpages":[["03-ingestion/web-clipper",7],["03-ingestion/chat-exports",5],["10-troubleshooting/broken-links",5],["03-ingestion/youtube-transcripts",4],["10-troubleshooting/common-failures",4],["03-ingestion/newsletters-and-email",6],["06-agents/scheduled-maintenance",5],["01-concepts/wiki-vs-rag",4],["03-ingestion/pdfs-and-books",4],["09-maintenance/privacy-and-secrets",4],["06-agents/agent-roles",3],["06-agents/subagents",3],["07-retrieval/rag-on-top",3],["09-maintenance/lint-and-health",3],["10-troubleshooting/faq",3],["track-graph/why-graphs",2],["track-graph/build-use",2],["01-concepts/two-layers",2],["02-setup/claude-code-setup",2],["02-setup/git-and-sync",2],["03-ingestion/voice-and-meetings",2],["03-ingestion/bulk-backfill",2],["04-structuring/contradictions-and-supersession",2],["05-graphs/typed-links",2],["05-graphs/metrics",2],["06-agents/hooks",2],["06-agents/skills-and-commands",2],["06-agents/safety-and-guardrails",2],["08-outputs/research-reports",2]],"total":91,"val":2.72},{"id":"term:wikilink","type":"term","group":"term","label":"Wikilink","desc":"A [[Page name]] link between notes. The cheapest possible graph edge, and the reason the vault compounds.","inpages":[["track-graph/graph-stores",2],["01-concepts/why-markdown-and-plain-text",1],["track-graph/build-extract",2],["05-graphs/graph-basics",2],["05-graphs/typed-links",2],["09-maintenance/backups-and-portability",2]],"total":11,"val":1.83},{"id":"term:backlinks","type":"term","group":"term","label":"Backlinks","desc":"The links pointing at a page. How you find everything that leans on an idea without searching for it.","inpages":[["04-structuring/naming-and-aliases",2],["04-structuring/dedupe-and-merge",2],["04-structuring/linking-rules",1],["05-graphs/metrics",1]],"total":6,"val":1.7},{"id":"term:atomic-pages","type":"term","group":"term","label":"Atomic pages","desc":"One idea per page, so a link means one thing and a page can be reused across every context it belongs to.","inpages":[["04-structuring/atomic-pages",2],["01-concepts/pkm-lineage",2],["01-concepts/llm-wiki-pattern",1],["05-graphs/graph-vs-vectors",1]],"total":6,"val":1.7},{"id":"term:page-types","type":"term","group":"term","label":"Page types","desc":"Source, concept, entity, synthesis: each with its own contract for what goes in it and how it links.","inpages":[["04-structuring/page-types",6],["04-structuring/atomic-pages",5],["03-ingestion/pdfs-and-books",3],["03-ingestion/voice-and-meetings",3],["03-ingestion/newsletters-and-email",3],["04-structuring/frontmatter-schema",3],["08-outputs/writing-from-the-vault",3],["10-troubleshooting/agent-writes-garbage",3]],"total":29,"val":1.95},{"id":"term:frontmatter","type":"term","group":"term","label":"Frontmatter","desc":"The YAML block at the top of a page: type, sources, dates, status. What makes the vault queryable without a database.","inpages":[["04-structuring/frontmatter-schema",1],["course-5-evals/evals-practice",6],["02-setup/obsidian-plugins",3],["03-ingestion/bulk-backfill",2],["05-graphs/typed-links",2]],"total":14,"val":1.77},{"id":"term:aliases","type":"term","group":"term","label":"Aliases","desc":"Other names a page answers to, so links and searches land on one page instead of spawning duplicates.","inpages":[["04-structuring/naming-and-aliases",10],["04-structuring/dedupe-and-merge",5],["04-structuring/frontmatter-schema",4],["07-retrieval/rag-on-top",2],["10-troubleshooting/broken-links",2]],"total":23,"val":1.77},{"id":"term:contradictions","type":"term","group":"term","label":"Contradictions","desc":"When a new source disagrees, both positions stay, dated and sourced. The record of what you believed and why.","inpages":[["04-structuring/contradictions-and-supersession",6],["07-retrieval/query-patterns",3],["05-graphs/typed-links",6],["track-graph/build-extract",2],["04-structuring/dedupe-and-merge",2],["06-agents/skills-and-commands",2],["09-maintenance/review-cadence",2]],"total":23,"val":1.89},{"id":"term:index-log","type":"term","group":"term","label":"index.md and log.md","desc":"The catalog the agent reads first, and one line per run so every change to the vault is auditable.","inpages":[["04-structuring/index-and-log",2],["03-ingestion/bulk-backfill",3],["07-retrieval/asking-questions",3],["02-setup/claude-code-setup",2],["02-setup/vault-structure",2],["04-structuring/contradictions-and-supersession",2],["06-agents/scheduled-maintenance",2],["06-agents/subagents",2],["09-maintenance/lint-and-health",2],["10-troubleshooting/broken-links",2]],"total":22,"val":2.05},{"id":"term:knowledge-graph","type":"term","group":"term","label":"Knowledge graph","desc":"Pages as nodes, links as edges. In this setup the graph is the filesystem: no server, no separate store.","inpages":[["track-graph/why-graphs",2],["05-graphs/obsidian-graph-view",0],["track-graph/resources",5],["track-graph/tools",3],["05-graphs/graph-basics",2],["05-graphs/graphrag",2],["05-graphs/exporting-your-graph",2],["10-troubleshooting/faq",2]],"total":18,"val":1.95},{"id":"term:typed-links","type":"term","group":"term","label":"Typed links","desc":"A link that says how two pages relate, such as supports, contradicts or extends. Only typed when the source says so.","inpages":[["05-graphs/typed-links",1],["track-graph/why-graphs",2],["track-graph/graph-stores",1],["05-graphs/obsidian-graph-view",1]],"total":5,"val":1.7},{"id":"term:orphans","type":"term","group":"term","label":"Orphans","desc":"Pages nothing links to. Invisible within a week, and the first number to watch when a vault is decaying.","inpages":[["05-graphs/metrics",4],["10-troubleshooting/broken-links",3],["track-graph/build-use",2],["05-graphs/obsidian-graph-view",2],["10-troubleshooting/vault-too-big",2]],"total":13,"val":1.77},{"id":"term:hubs","type":"term","group":"term","label":"Hubs and bridges","desc":"Hubs are pages everything links to; bridges connect clusters that would otherwise not touch. Both tell you where the thinking is.","inpages":[["04-structuring/linking-rules",4],["05-graphs/exporting-your-graph",4],["05-graphs/graph-basics",5],["05-graphs/obsidian-graph-view",3],["track-graph/build-use",2],["06-agents/agent-roles",2],["10-troubleshooting/common-failures",2]],"total":22,"val":1.89},{"id":"term:clusters","type":"term","group":"term","label":"Clusters","desc":"Groups of densely linked pages. The ones you did not plan are where the vault is telling you something.","inpages":[["track-evals/build-traces",3],["05-graphs/graphrag",10],["track-graph/build-query",8],["track-graph/graphrag",7],["track-graph/build-use",6],["05-graphs/obsidian-graph-view",5],["track-graph/tools",3],["02-setup/mcp-obsidian",3],["05-graphs/graph-basics",3],["track-graph/graph-stores",2]],"total":50,"val":2.05},{"id":"term:graphrag","type":"term","group":"term","label":"GraphRAG","desc":"Retrieval over an extracted entity graph with community summaries. Built for global questions about a whole corpus.","inpages":[["track-graph/graphrag",4],["05-graphs/graphrag",2],["track-graph/resources",6],["track-graph/tools",4],["track-graph/build-query",4],["track-graph/graph-stores",2]],"total":22,"val":1.83},{"id":"term:rag","type":"term","group":"term","label":"RAG","desc":"Retrieval-augmented generation: fetch chunks by similarity at question time. Nothing accumulates between questions.","inpages":[["01-concepts/wiki-vs-rag",8],["track-graph/why-graphs",1],["07-retrieval/rag-on-top",0],["track-graph/build-extract",6],["track-graph/build-query",5],["track-graph/resources",5],["04-structuring/naming-and-aliases",5],["track-graph/graphrag",3]],"total":33,"val":1.95},{"id":"term:vectors","type":"term","group":"term","label":"Embeddings and vectors","desc":"Similarity search over chunk embeddings. Worth adding only once a structured, linked vault stops answering.","inpages":[["05-graphs/graph-vs-vectors",3],["track-graph/why-graphs",6],["07-retrieval/rag-on-top",6],["01-concepts/wiki-vs-rag",3]],"total":18,"val":1.7},{"id":"term:context-window","type":"term","group":"term","label":"Context window","desc":"What the model can see at once. Everything the agent knows in a session competes for this budget.","inpages":[["track-harness/build-guardrails",3],["track-harness/what-a-harness-is",2],["07-retrieval/context-budget",0],["07-retrieval/rag-on-top",2]],"total":7,"val":1.7},{"id":"term:lost-in-middle","type":"term","group":"term","label":"Lost in the middle","desc":"Models recall the start and end of a long context far better than the middle. Why stuffing everything in fails.","inpages":[["track-graph/build-query",4],["track-graph/build-extract",2],["course-1-context/how-models-read",1],["track-graph/build-use",1]],"total":8,"val":1.7},{"id":"term:skills","type":"term","group":"term","label":"Skills","desc":"A folder with a SKILL.md: instructions the agent loads only when the task matches. One per workflow.","inpages":[["06-agents/skills-and-commands",16],["track-harness/claude-code-as-harness",4],["track-evals/tooling",2],["10-troubleshooting/faq",2]],"total":24,"val":1.7},{"id":"term:slash-commands","type":"term","group":"term","label":"Slash commands","desc":"Short, scoped entry points into a skill, typed as /name. The handle you actually use day to day.","inpages":[["track-graph/build-use",2],["06-agents/skills-and-commands",1],["03-ingestion/web-clipper",1],["03-ingestion/bulk-backfill",1]],"total":5,"val":1.7},{"id":"term:subagents","type":"term","group":"term","label":"Subagents","desc":"Separate agents with their own context and a narrow tool list. Most of them here are read-only by design.","inpages":[["06-agents/subagents",9],["course-1-context/context-practice",4],["track-harness/claude-code-as-harness",3],["07-retrieval/context-budget",2]],"total":18,"val":1.7},{"id":"term:hooks","type":"term","group":"term","label":"Hooks","desc":"Code the harness runs on events, before a tool call or after an edit. Rules the model cannot forget.","inpages":[["06-agents/hooks",12],["course-4-harness/harness-practice",10],["track-harness/claude-code-as-harness",8],["course-4-harness/the-four-rings",6],["track-harness/build-graduate",3],["track-harness/context-engineering",2]],"total":41,"val":1.83},{"id":"term:scheduled","type":"term","group":"term","label":"Scheduled maintenance","desc":"Lint, link and review runs on a timer, so the vault files itself while you sleep.","inpages":[["track-loop/build-overnight",7],["track-loop/patterns",2],["course-5-evals/evals-practice",1],["track-evals/build-ci",9],["06-agents/scheduled-maintenance",7],["course-2-loop/loop-practice",4],["course-6-production/from-prototype",4],["02-setup/claude-code-setup",4],["06-agents/hooks",3],["06-agents/skills-and-commands",3],["06-agents/safety-and-guardrails",3]],"total":47,"val":2.09},{"id":"term:lint","type":"term","group":"term","label":"Lint and health","desc":"Mechanical checks for broken links, orphans, missing frontmatter and stale pages. The vault's test suite.","inpages":[["09-maintenance/lint-and-health",3],["track-graph/build-use",2],["01-concepts/llm-wiki-pattern",3],["course-2-loop/the-four-parts",2],["03-ingestion/bulk-backfill",2],["10-troubleshooting/broken-links",2]],"total":14,"val":1.83},{"id":"term:git","type":"term","group":"term","label":"Git","desc":"Version history for the vault. Every agent run becomes reviewable and every bad run reversible.","inpages":[["02-setup/git-and-sync",13],["09-maintenance/versioning-with-git",13],["06-agents/safety-and-guardrails",5],["01-concepts/why-markdown-and-plain-text",1],["track-evals/build-ci",10],["track-loop/build-overnight",9],["course-4-harness/harness-practice",6],["06-agents/scheduled-maintenance",6],["track-harness/build-guardrails",5],["02-setup/obsidian-plugins",3],["course-4-harness/the-four-rings",2],["track-loop/build-critic",2],["04-structuring/dedupe-and-merge",2],["06-agents/hooks",2],["09-maintenance/lint-and-health",2],["09-maintenance/backups-and-portability",2]],"total":83,"val":2.3},{"id":"term:privacy","type":"term","group":"term","label":"Privacy and secrets","desc":"What must never reach a page, a prompt or a public export, and the checks that keep it out.","inpages":[["03-ingestion/chat-exports",3],["09-maintenance/privacy-and-secrets",0],["track-evals/build-ci",2],["02-setup/git-and-sync",2],["06-agents/safety-and-guardrails",2]],"total":9,"val":1.77},{"id":"term:web-clipper","type":"term","group":"term","label":"Web Clipper","desc":"Obsidian's browser extension. One click turns an article into a markdown file in raw/.","inpages":[["03-ingestion/web-clipper",10],["03-ingestion/newsletters-and-email",1],["03-ingestion/bulk-backfill",1],["06-agents/scheduled-maintenance",1]],"total":13,"val":1.7},{"id":"term:transcripts","type":"term","group":"term","label":"Transcripts","desc":"Video, voice and meetings turned into text so they can be ingested like anything else.","inpages":[["03-ingestion/youtube-transcripts",10],["03-ingestion/voice-and-meetings",4],["track-harness/context-engineering",3],["track-harness/build-graduate",3],["track-loop/context-hygiene",3],["course-0-map/agents-or-workflows",2],["course-1-context/the-four-places",2],["course-4-harness/harness-practice",2],["course-5-evals/two-kinds-of-checks",2],["track-harness/build-guardrails",2],["track-harness/resources",2],["10-troubleshooting/agent-writes-garbage",2]],"total":37,"val":2.14},{"id":"term:agent","type":"term","group":"term","label":"Agent","desc":"A model in a loop with tools, deciding its own next step until a goal is met or a stop rule fires.","inpages":[["course-0-map/what-an-agent-is",13],["course-0-map/agents-or-workflows",11],["course-6-production/operating-agents",9],["course-6-production/day-one-plan",9],["course-6-production/from-prototype",7],["track-evals/agent-evals",7],["01-concepts/why-markdown-and-plain-text",5],["02-setup/git-and-sync",5],["10-troubleshooting/common-failures",3],["track-jev/jev-in-an-agent-stack",2],["06-agents/agent-roles",2],["10-troubleshooting/agent-writes-garbage",2],["course-4-harness/harness-practice",15],["track-harness/resources",14],["course-5-evals/two-kinds-of-checks",11],["track-loop/what-loop-engineering-is",11],["course-2-loop/loop-practice",9],["track-harness/harness-landscape",9],["track-loop/build-goal-test",9],["course-4-harness/the-office",8],["course-4-harness/the-four-rings",8],["track-harness/context-engineering",8],["track-loop/resources",8],["02-setup/claude-md",8],["course-2-loop/loops-vs-workflows",7],["track-harness/claude-code-as-harness",7],["02-setup/claude-code-setup",7],["06-agents/safety-and-guardrails",7],["course-0-map/five-layers",6],["track-graph/build-use",6],["track-harness/what-a-harness-is",6],["track-harness/build-the-loop",6],["track-harness/build-graduate",6],["track-loop/stop-conditions",6],["04-structuring/index-and-log",6],["track-harness/tools-and-mcp",5],["track-loop/critics-and-verification",5],["track-loop/context-hygiene",5],["03-ingestion/web-clipper",5],["07-retrieval/search-tools",5],["course-1-context/context-practice",4],["course-5-evals/judges-and-golden-sets",4],["track-graph/tools",4],["track-harness/build-guardrails",4],["01-concepts/llm-wiki-pattern",4],["02-setup/obsidian-install-and-vault",4],["04-structuring/dedupe-and-merge",4],["07-retrieval/context-budget",4],["10-troubleshooting/faq",4],["course-3-gate/cheap-decisions",3],["course-5-evals/evals-practice",3],["track-graph/why-graphs",3],["track-graph/graph-stores",3],["track-graph/build-extract",3],["track-graph/resources",3],["track-jev/what-jev-is-good-for",3],["track-jev/build-router",3],["track-evals/why-evals",3],["track-evals/tooling",3],["track-evals/resources",3],["01-concepts/pkm-lineage",3],["02-setup/project-scoping",3],["02-setup/live-data",3],["02-setup/obsidian-plugins",3],["03-ingestion/youtube-transcripts",3],["03-ingestion/pdfs-and-books",3],["03-ingestion/chat-exports",3],["04-structuring/atomic-pages",3],["04-structuring/frontmatter-schema",3],["04-structuring/naming-and-aliases",3],["05-graphs/graph-basics",3],["05-graphs/typed-links",3],["07-retrieval/asking-questions",3],["10-troubleshooting/broken-links",3],["10-troubleshooting/vault-too-big",3],["course-1-context/how-models-read",2],["course-1-context/the-four-places",2],["course-3-gate/gate-practice",2],["track-graph/building-graphs-with-llms",2],["track-jev/resources",2],["track-loop/build-overnight",2],["track-evals/designing-evals",2],["02-setup/mcp-obsidian",2],["02-setup/vault-structure",2],["03-ingestion/bulk-backfill",2],["04-structuring/page-types",2],["06-agents/subagents",2],["06-agents/skills-and-commands",2],["09-maintenance/versioning-with-git",2]],"total":430,"val":3.93},{"id":"term:five-layers","type":"term","group":"term","label":"Five layers","desc":"The course's frame for everything around the model: context, loop, gate, harness, evals.","inpages":[["course-0-map/five-layers",2],["course-0-map/agents-or-workflows",2],["course-6-production/day-one-plan",2],["course-0-map/what-an-agent-is",1]],"total":7,"val":1.7},{"id":"term:workflows","type":"term","group":"term","label":"Workflows vs agents","desc":"Code decides the next step in a workflow; the model decides in an agent. Most production systems are a hybrid.","inpages":[["course-0-map/agents-or-workflows",7],["course-2-loop/loops-vs-workflows",7],["course-0-map/what-an-agent-is",3],["track-evals/build-ci",2],["course-0-map/five-layers",2],["track-jev/what-jev-is-good-for",2],["06-agents/skills-and-commands",2]],"total":25,"val":1.89},{"id":"term:context-engineering","type":"term","group":"term","label":"Context engineering","desc":"Deciding what goes into the system prompt, tools, history and tail, and in what order, so the model sees the right things.","inpages":[["course-1-context/context-practice",3],["track-harness/context-engineering",2],["course-1-context/the-four-places",1],["track-harness/resources",4],["course-1-context/how-models-read",2],["track-loop/resources",2]],"total":14,"val":1.83},{"id":"term:caching","type":"term","group":"term","label":"Prompt caching","desc":"Reusing the stable prefix of a prompt across calls. Cheaper and faster, if the context is laid out for it.","inpages":[["course-1-context/how-models-read",11],["course-1-context/the-four-places",6],["course-1-context/context-practice",4],["course-6-production/operating-agents",3],["course-6-production/from-prototype",2],["track-harness/resources",2]],"total":28,"val":1.83},{"id":"term:memory","type":"term","group":"term","label":"Memory","desc":"What persists between sessions. Here it is files the agent reads back, not a hidden feature of the model.","inpages":[["track-graph/tools",6],["track-harness/context-engineering",6],["track-loop/context-hygiene",4],["track-harness/claude-code-as-harness",1],["track-graph/resources",5],["course-1-context/the-four-places",4],["track-graph/why-graphs",3],["course-1-context/context-practice",2],["course-4-harness/the-office",2],["track-graph/building-graphs-with-llms",2],["track-graph/graph-stores",2],["10-troubleshooting/faq",2]],"total":39,"val":2.14},{"id":"term:loop","type":"term","group":"term","label":"Agent loop","desc":"Goal, act, check, repeat. The loop is the control system around the model, and most failures live in it.","inpages":[["track-loop/what-loop-engineering-is",18],["course-2-loop/loop-practice",15],["track-loop/patterns",12],["track-loop/context-hygiene",10],["course-2-loop/loops-vs-workflows",9],["course-2-loop/the-four-parts",9],["course-0-map/five-layers",7],["course-0-map/agents-or-workflows",6],["course-6-production/operating-agents",4],["track-jev/jev-in-an-agent-stack",4],["track-harness/what-a-harness-is",4],["track-loop/build-overnight",4],["course-6-production/day-one-plan",3],["track-graph/build-use",2],["track-graph/building-graphs-with-llms",1],["track-harness/build-the-loop",1],["08-outputs/writing-from-the-vault",1],["track-loop/resources",17],["track-loop/stop-conditions",8],["track-loop/critics-and-verification",7],["course-0-map/what-an-agent-is",4],["track-harness/build-graduate",4],["course-1-context/the-four-places",3],["course-4-harness/the-office",3],["course-4-harness/harness-practice",3],["track-harness/harness-landscape",3],["track-loop/build-goal-test",3],["course-1-context/how-models-read",2],["course-6-production/from-prototype",2],["track-harness/resources",2]],"total":171,"val":2.74},{"id":"term:stop-conditions","type":"term","group":"term","label":"Stop conditions","desc":"When a loop must end: goal met, budget spent, no progress. A loop without one runs until it does damage.","inpages":[["course-2-loop/the-four-parts",11],["track-harness/build-guardrails",6],["track-loop/stop-conditions",4],["course-5-evals/two-kinds-of-checks",2],["track-harness/context-engineering",2],["07-retrieval/context-budget",1],["track-loop/what-loop-engineering-is",5],["course-2-loop/loop-practice",3],["track-loop/build-overnight",3],["course-2-loop/loops-vs-workflows",2],["course-6-production/day-one-plan",2],["track-harness/build-graduate",2],["track-loop/critics-and-verification",2],["07-retrieval/rag-on-top",2]],"total":47,"val":2.22},{"id":"term:critics","type":"term","group":"term","label":"Critics and checkers","desc":"A check outside the model that judges the result: tests, a compiler, a separate reviewer. Self-review is not one.","inpages":[["track-loop/critics-and-verification",19],["track-loop/build-critic",18],["course-2-loop/the-four-parts",9],["track-loop/patterns",12],["course-2-loop/loop-practice",6],["track-loop/resources",5],["track-loop/what-loop-engineering-is",4],["track-loop/context-hygiene",4],["track-loop/build-goal-test",3],["track-loop/stop-conditions",2]],"total":82,"val":2.05},{"id":"term:gate","type":"term","group":"term","label":"The gate","desc":"A cheap, narrow decision in front of an expensive model: one bounded question per item, route on the answer.","inpages":[["course-3-gate/gate-practice",21],["course-3-gate/cheap-decisions",15],["course-3-gate/gates-in-practice",11],["course-0-map/five-layers",5],["track-evals/build-ci",5],["track-jev/build-router",3],["track-harness/build-guardrails",3],["course-6-production/day-one-plan",2],["track-jev/jev-in-an-agent-stack",3],["track-jev/build-swap-in-jev",3],["course-6-production/from-prototype",2],["course-6-production/operating-agents",2],["track-jev/system-one-models",2],["track-jev/build-decision-endpoint",2],["track-harness/tools-and-mcp",2],["track-harness/build-graduate",2],["track-evals/agent-evals",2],["track-evals/tooling",2]],"total":87,"val":2.37},{"id":"term:system-one","type":"term","group":"term","label":"System One models","desc":"Small fast models that return typed decisions with confidence instead of text. Jev is one way to build the gate.","inpages":[["track-jev/resources",15],["track-jev/build-swap-in-jev",12],["track-jev/jev-in-an-agent-stack",11],["course-3-gate/gates-in-practice",9],["track-jev/what-jev-is-good-for",8],["track-jev/getting-started",8],["track-jev/system-one-models",6],["track-jev/build-decision-endpoint",6],["course-3-gate/cheap-decisions",5],["track-jev/build-router",4],["course-3-gate/gate-practice",2]],"total":86,"val":2.09},{"id":"term:harness","type":"term","group":"term","label":"Harness","desc":"The office around the model: containment, guides, sensors and permissions, built from the outside in.","inpages":[["track-harness/what-a-harness-is",14],["course-4-harness/the-office",13],["track-harness/harness-landscape",10],["course-4-harness/harness-practice",9],["course-0-map/five-layers",5],["track-loop/what-loop-engineering-is",5],["course-6-production/day-one-plan",4],["track-harness/claude-code-as-harness",4],["track-evals/agent-evals",2],["course-4-harness/the-four-rings",5],["track-harness/resources",5],["track-harness/build-guardrails",4],["track-harness/context-engineering",3],["track-harness/build-graduate",3],["course-0-map/agents-or-workflows",2],["course-1-context/context-practice",2],["course-6-production/operating-agents",2],["track-harness/build-the-loop",2],["track-loop/context-hygiene",2],["track-loop/build-goal-test",2]],"total":98,"val":2.44},{"id":"term:guardrails","type":"term","group":"term","label":"Guardrails and permissions","desc":"What the agent is allowed to touch, what it must ask about, and the walls that hold when it is wrong.","inpages":[["course-4-harness/the-four-rings",7],["course-4-harness/harness-practice",5],["06-agents/safety-and-guardrails",4],["track-harness/what-a-harness-is",3],["track-harness/claude-code-as-harness",3],["track-harness/build-guardrails",3],["02-setup/claude-code-setup",2],["course-6-production/operating-agents",3],["course-6-production/day-one-plan",3],["track-harness/build-graduate",3],["track-harness/harness-landscape",2],["02-setup/live-data",2]],"total":40,"val":2.14},{"id":"term:tools-tool-use","type":"term","group":"term","label":"Tool use","desc":"The model asking the harness to run something on its behalf. Every capability an agent has arrives this way.","inpages":[["track-harness/build-the-loop",5],["track-harness/build-guardrails",4],["course-6-production/operating-agents",2],["track-jev/jev-in-an-agent-stack",2],["track-harness/what-a-harness-is",2],["track-harness/build-graduate",2]],"total":17,"val":1.83},{"id":"term:evals","type":"term","group":"term","label":"Evals","desc":"The same test, every month: checks on real behaviour so a prompt or model change cannot quietly break things.","inpages":[["track-evals/tooling",19],["track-evals/resources",15],["course-0-map/five-layers",7],["track-evals/why-evals",7],["track-evals/agent-evals",5],["track-evals/designing-evals",3],["course-5-evals/evals-practice",2],["course-6-production/day-one-plan",2],["track-evals/build-ci",9],["course-5-evals/two-kinds-of-checks",4],["track-jev/resources",4],["course-5-evals/judges-and-golden-sets",3],["course-0-map/agents-or-workflows",2],["course-6-production/from-prototype",2],["course-6-production/operating-agents",2],["track-jev/what-jev-is-good-for",2],["track-evals/llm-as-judge",2],["track-evals/build-traces",2]],"total":92,"val":2.37},{"id":"term:golden-set","type":"term","group":"term","label":"Golden set","desc":"The fixed collection of cases every eval runs against, including the failures you have already seen.","inpages":[["course-5-evals/judges-and-golden-sets",6],["track-evals/designing-evals",4],["track-evals/build-traces",4],["track-evals/build-suite",3],["track-evals/build-ci",3]],"total":20,"val":1.77},{"id":"term:llm-judge","type":"term","group":"term","label":"LLM as judge","desc":"A model grading outputs against a rubric. Useful once the judge itself has been checked against human labels.","inpages":[["track-evals/llm-as-judge",18],["course-5-evals/judges-and-golden-sets",13],["track-evals/build-suite",7],["course-5-evals/evals-practice",6],["track-evals/resources",5],["course-3-gate/gate-practice",4],["track-loop/build-critic",4],["course-5-evals/two-kinds-of-checks",3],["track-loop/critics-and-verification",3],["course-2-loop/the-four-parts",2],["track-evals/agent-evals",2],["track-evals/build-ci",2]],"total":69,"val":2.14},{"id":"term:traces","type":"term","group":"term","label":"Traces","desc":"The full record of what an agent did, step by step. What behavioural evals and debugging both read.","inpages":[["course-5-evals/evals-practice",17],["track-evals/build-traces",10],["course-5-evals/two-kinds-of-checks",7],["course-6-production/operating-agents",7],["track-evals/agent-evals",4],["track-evals/tooling",5],["track-evals/why-evals",3],["track-evals/resources",3],["course-6-production/from-prototype",2]],"total":58,"val":2.0},{"id":"term:production","type":"term","group":"term","label":"Production","desc":"The step from a demo that works once to a system with tracing, cost control, security and an owner.","inpages":[["course-6-production/from-prototype",9],["course-5-evals/judges-and-golden-sets",4],["course-6-production/operating-agents",5],["track-harness/resources",5],["course-2-loop/loops-vs-workflows",3],["course-0-map/agents-or-workflows",2],["course-2-loop/the-four-parts",2],["course-2-loop/loop-practice",2],["track-jev/getting-started",2],["track-harness/harness-landscape",2]],"total":36,"val":2.05}],"edges":[{"s":"root","t":"col:course","k":"contains"},{"s":"root","t":"col:handbook","k":"contains"},{"s":"root","t":"col:guide","k":"contains"},{"s":"root","t":"col:glossary","k":"contains"},{"s":"root","t":"col:toolkit","k":"contains"},{"s":"root","t":"col:resources","k":"contains"},{"s":"col:course","t":"sec:course-0-map","k":"contains"},{"s":"sec:course-0-map","t":"course-0-map/what-an-agent-is","k":"contains"},{"s":"sec:course-0-map","t":"course-0-map/five-layers","k":"contains"},{"s":"sec:course-0-map","t":"course-0-map/agents-or-workflows","k":"contains"},{"s":"col:course","t":"sec:course-1-context","k":"contains"},{"s":"sec:course-1-context","t":"course-1-context/how-models-read","k":"contains"},{"s":"sec:course-1-context","t":"course-1-context/the-four-places","k":"contains"},{"s":"sec:course-1-context","t":"course-1-context/context-practice","k":"contains"},{"s":"col:course","t":"sec:course-2-loop","k":"contains"},{"s":"sec:course-2-loop","t":"course-2-loop/loops-vs-workflows","k":"contains"},{"s":"sec:course-2-loop","t":"course-2-loop/the-four-parts","k":"contains"},{"s":"sec:course-2-loop","t":"course-2-loop/loop-practice","k":"contains"},{"s":"col:course","t":"sec:course-3-gate","k":"contains"},{"s":"sec:course-3-gate","t":"course-3-gate/cheap-decisions","k":"contains"},{"s":"sec:course-3-gate","t":"course-3-gate/gates-in-practice","k":"contains"},{"s":"sec:course-3-gate","t":"course-3-gate/gate-practice","k":"contains"},{"s":"col:course","t":"sec:course-4-harness","k":"contains"},{"s":"sec:course-4-harness","t":"course-4-harness/the-office","k":"contains"},{"s":"sec:course-4-harness","t":"course-4-harness/the-four-rings","k":"contains"},{"s":"sec:course-4-harness","t":"course-4-harness/harness-practice","k":"contains"},{"s":"col:course","t":"sec:course-5-evals","k":"contains"},{"s":"sec:course-5-evals","t":"course-5-evals/two-kinds-of-checks","k":"contains"},{"s":"sec:course-5-evals","t":"course-5-evals/judges-and-golden-sets","k":"contains"},{"s":"sec:course-5-evals","t":"course-5-evals/evals-practice","k":"contains"},{"s":"col:course","t":"sec:course-6-production","k":"contains"},{"s":"sec:course-6-production","t":"course-6-production/from-prototype","k":"contains"},{"s":"sec:course-6-production","t":"course-6-production/operating-agents","k":"contains"},{"s":"sec:course-6-production","t":"course-6-production/day-one-plan","k":"contains"},{"s":"col:handbook","t":"sec:track-graph","k":"contains"},{"s":"sec:track-graph","t":"track-graph/why-graphs","k":"contains"},{"s":"sec:track-graph","t":"track-graph/graphrag","k":"contains"},{"s":"sec:track-graph","t":"track-graph/building-graphs-with-llms","k":"contains"},{"s":"sec:track-graph","t":"track-graph/graph-stores","k":"contains"},{"s":"sec:track-graph","t":"track-graph/tools","k":"contains"},{"s":"sec:track-graph","t":"track-graph/build-extract","k":"contains"},{"s":"sec:track-graph","t":"track-graph/build-query","k":"contains"},{"s":"sec:track-graph","t":"track-graph/build-use","k":"contains"},{"s":"sec:track-graph","t":"track-graph/resources","k":"contains"},{"s":"col:handbook","t":"sec:track-jev","k":"contains"},{"s":"sec:track-jev","t":"track-jev/system-one-models","k":"contains"},{"s":"sec:track-jev","t":"track-jev/what-jev-is-good-for","k":"contains"},{"s":"sec:track-jev","t":"track-jev/jev-in-an-agent-stack","k":"contains"},{"s":"sec:track-jev","t":"track-jev/getting-started","k":"contains"},{"s":"sec:track-jev","t":"track-jev/build-decision-endpoint","k":"contains"},{"s":"sec:track-jev","t":"track-jev/build-router","k":"contains"},{"s":"sec:track-jev","t":"track-jev/build-swap-in-jev","k":"contains"},{"s":"sec:track-jev","t":"track-jev/resources","k":"contains"},{"s":"col:handbook","t":"sec:track-harness","k":"contains"},{"s":"sec:track-harness","t":"track-harness/what-a-harness-is","k":"contains"},{"s":"sec:track-harness","t":"track-harness/claude-code-as-harness","k":"contains"},{"s":"sec:track-harness","t":"track-harness/context-engineering","k":"contains"},{"s":"sec:track-harness","t":"track-harness/tools-and-mcp","k":"contains"},{"s":"sec:track-harness","t":"track-harness/harness-landscape","k":"contains"},{"s":"sec:track-harness","t":"track-harness/build-the-loop","k":"contains"},{"s":"sec:track-harness","t":"track-harness/build-guardrails","k":"contains"},{"s":"sec:track-harness","t":"track-harness/build-graduate","k":"contains"},{"s":"sec:track-harness","t":"track-harness/resources","k":"contains"},{"s":"col:handbook","t":"sec:track-loop","k":"contains"},{"s":"sec:track-loop","t":"track-loop/what-loop-engineering-is","k":"contains"},{"s":"sec:track-loop","t":"track-loop/stop-conditions","k":"contains"},{"s":"sec:track-loop","t":"track-loop/critics-and-verification","k":"contains"},{"s":"sec:track-loop","t":"track-loop/context-hygiene","k":"contains"},{"s":"sec:track-loop","t":"track-loop/patterns","k":"contains"},{"s":"sec:track-loop","t":"track-loop/build-goal-test","k":"contains"},{"s":"sec:track-loop","t":"track-loop/build-critic","k":"contains"},{"s":"sec:track-loop","t":"track-loop/build-overnight","k":"contains"},{"s":"sec:track-loop","t":"track-loop/resources","k":"contains"},{"s":"col:handbook","t":"sec:track-evals","k":"contains"},{"s":"sec:track-evals","t":"track-evals/why-evals","k":"contains"},{"s":"sec:track-evals","t":"track-evals/designing-evals","k":"contains"},{"s":"sec:track-evals","t":"track-evals/llm-as-judge","k":"contains"},{"s":"sec:track-evals","t":"track-evals/agent-evals","k":"contains"},{"s":"sec:track-evals","t":"track-evals/tooling","k":"contains"},{"s":"sec:track-evals","t":"track-evals/build-traces","k":"contains"},{"s":"sec:track-evals","t":"track-evals/build-suite","k":"contains"},{"s":"sec:track-evals","t":"track-evals/build-ci","k":"contains"},{"s":"sec:track-evals","t":"track-evals/resources","k":"contains"},{"s":"col:guide","t":"sec:01-concepts","k":"contains"},{"s":"sec:01-concepts","t":"01-concepts/what-is-a-second-brain","k":"contains"},{"s":"sec:01-concepts","t":"01-concepts/the-save-for-later-paradox","k":"contains"},{"s":"sec:01-concepts","t":"01-concepts/llm-wiki-pattern","k":"contains"},{"s":"sec:01-concepts","t":"01-concepts/two-layers","k":"contains"},{"s":"sec:01-concepts","t":"01-concepts/why-markdown-and-plain-text","k":"contains"},{"s":"sec:01-concepts","t":"01-concepts/pkm-lineage","k":"contains"},{"s":"sec:01-concepts","t":"01-concepts/wiki-vs-rag","k":"contains"},{"s":"sec:01-concepts","t":"01-concepts/what-good-looks-like","k":"contains"},{"s":"col:guide","t":"sec:02-setup","k":"contains"},{"s":"sec:02-setup","t":"02-setup/obsidian-install-and-vault","k":"contains"},{"s":"sec:02-setup","t":"02-setup/claude-code-setup","k":"contains"},{"s":"sec:02-setup","t":"02-setup/mcp-obsidian","k":"contains"},{"s":"sec:02-setup","t":"02-setup/claude-md","k":"contains"},{"s":"sec:02-setup","t":"02-setup/vault-structure","k":"contains"},{"s":"sec:02-setup","t":"02-setup/project-scoping","k":"contains"},{"s":"sec:02-setup","t":"02-setup/live-data","k":"contains"},{"s":"sec:02-setup","t":"02-setup/obsidian-plugins","k":"contains"},{"s":"sec:02-setup","t":"02-setup/git-and-sync","k":"contains"},{"s":"col:guide","t":"sec:03-ingestion","k":"contains"},{"s":"sec:03-ingestion","t":"03-ingestion/web-clipper","k":"contains"},{"s":"sec:03-ingestion","t":"03-ingestion/youtube-transcripts","k":"contains"},{"s":"sec:03-ingestion","t":"03-ingestion/pdfs-and-books","k":"contains"},{"s":"sec:03-ingestion","t":"03-ingestion/chat-exports","k":"contains"},{"s":"sec:03-ingestion","t":"03-ingestion/voice-and-meetings","k":"contains"},{"s":"sec:03-ingestion","t":"03-ingestion/newsletters-and-email","k":"contains"},{"s":"sec:03-ingestion","t":"03-ingestion/bulk-backfill","k":"contains"},{"s":"col:guide","t":"sec:04-structuring","k":"contains"},{"s":"sec:04-structuring","t":"04-structuring/atomic-pages","k":"contains"},{"s":"sec:04-structuring","t":"04-structuring/page-types","k":"contains"},{"s":"sec:04-structuring","t":"04-structuring/linking-rules","k":"contains"},{"s":"sec:04-structuring","t":"04-structuring/frontmatter-schema","k":"contains"},{"s":"sec:04-structuring","t":"04-structuring/naming-and-aliases","k":"contains"},{"s":"sec:04-structuring","t":"04-structuring/dedupe-and-merge","k":"contains"},{"s":"sec:04-structuring","t":"04-structuring/contradictions-and-supersession","k":"contains"},{"s":"sec:04-structuring","t":"04-structuring/index-and-log","k":"contains"},{"s":"col:guide","t":"sec:05-graphs","k":"contains"},{"s":"sec:05-graphs","t":"05-graphs/graph-basics","k":"contains"},{"s":"sec:05-graphs","t":"05-graphs/obsidian-graph-view","k":"contains"},{"s":"sec:05-graphs","t":"05-graphs/typed-links","k":"contains"},{"s":"sec:05-graphs","t":"05-graphs/graph-vs-vectors","k":"contains"},{"s":"sec:05-graphs","t":"05-graphs/graphrag","k":"contains"},{"s":"sec:05-graphs","t":"05-graphs/exporting-your-graph","k":"contains"},{"s":"sec:05-graphs","t":"05-graphs/metrics","k":"contains"},{"s":"col:guide","t":"sec:06-agents","k":"contains"},{"s":"sec:06-agents","t":"06-agents/agent-roles","k":"contains"},{"s":"sec:06-agents","t":"06-agents/scheduled-maintenance","k":"contains"},{"s":"sec:06-agents","t":"06-agents/subagents","k":"contains"},{"s":"sec:06-agents","t":"06-agents/hooks","k":"contains"},{"s":"sec:06-agents","t":"06-agents/skills-and-commands","k":"contains"},{"s":"sec:06-agents","t":"06-agents/safety-and-guardrails","k":"contains"},{"s":"col:guide","t":"sec:07-retrieval","k":"contains"},{"s":"sec:07-retrieval","t":"07-retrieval/asking-questions","k":"contains"},{"s":"sec:07-retrieval","t":"07-retrieval/query-patterns","k":"contains"},{"s":"sec:07-retrieval","t":"07-retrieval/search-tools","k":"contains"},{"s":"sec:07-retrieval","t":"07-retrieval/rag-on-top","k":"contains"},{"s":"sec:07-retrieval","t":"07-retrieval/context-budget","k":"contains"},{"s":"col:guide","t":"sec:08-outputs","k":"contains"},{"s":"sec:08-outputs","t":"08-outputs/writing-from-the-vault","k":"contains"},{"s":"sec:08-outputs","t":"08-outputs/research-reports","k":"contains"},{"s":"sec:08-outputs","t":"08-outputs/publishing-and-export","k":"contains"},{"s":"sec:08-outputs","t":"08-outputs/teaching-yourself","k":"contains"},{"s":"col:guide","t":"sec:09-maintenance","k":"contains"},{"s":"sec:09-maintenance","t":"09-maintenance/lint-and-health","k":"contains"},{"s":"sec:09-maintenance","t":"09-maintenance/review-cadence","k":"contains"},{"s":"sec:09-maintenance","t":"09-maintenance/versioning-with-git","k":"contains"},{"s":"sec:09-maintenance","t":"09-maintenance/backups-and-portability","k":"contains"},{"s":"sec:09-maintenance","t":"09-maintenance/privacy-and-secrets","k":"contains"},{"s":"sec:09-maintenance","t":"09-maintenance/scaling","k":"contains"},{"s":"col:guide","t":"sec:10-troubleshooting","k":"contains"},{"s":"sec:10-troubleshooting","t":"10-troubleshooting/common-failures","k":"contains"},{"s":"sec:10-troubleshooting","t":"10-troubleshooting/agent-writes-garbage","k":"contains"},{"s":"sec:10-troubleshooting","t":"10-troubleshooting/broken-links","k":"contains"},{"s":"sec:10-troubleshooting","t":"10-troubleshooting/vault-too-big","k":"contains"},{"s":"sec:10-troubleshooting","t":"10-troubleshooting/faq","k":"contains"},{"s":"sec:track-jev","t":"sec:course-3-gate","k":"handbook"},{"s":"sec:track-harness","t":"sec:course-4-harness","k":"handbook"},{"s":"sec:track-loop","t":"sec:course-2-loop","k":"handbook"},{"s":"sec:track-evals","t":"sec:course-5-evals","k":"handbook"},{"s":"col:toolkit","t":"hub:skills","k":"contains"},{"s":"col:toolkit","t":"hub:commands","k":"contains"},{"s":"col:toolkit","t":"hub:agents","k":"contains"},{"s":"col:toolkit","t":"hub:scripts","k":"contains"},{"s":"col:toolkit","t":"hub:plugin","k":"contains"},{"s":"col:toolkit","t":"hub:vault","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-backfill","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-changed-my-mind","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-chat-import","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-graph","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-ingest","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-lint","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-merge","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-metrics","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-privacy","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-project","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-publish","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-query","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-quiz","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-rename","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-report","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-review","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-transcript","k":"contains"},{"s":"hub:skills","t":"skill:second-brain-write","k":"contains"},{"s":"cmd:aliases","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:aliases","k":"contains","h":1},{"s":"cmd:archive","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:archive","k":"contains","h":1},{"s":"cmd:ask","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:ask","k":"contains","h":1},{"s":"cmd:audit","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:audit","k":"contains","h":1},{"s":"cmd:backfill","t":"skill:second-brain-backfill","k":"follows"},{"s":"hub:commands","t":"cmd:backfill","k":"contains","h":1},{"s":"cmd:bridges","t":"skill:second-brain-graph","k":"follows"},{"s":"hub:commands","t":"cmd:bridges","k":"contains","h":1},{"s":"cmd:changed-my-mind","t":"skill:second-brain-changed-my-mind","k":"follows"},{"s":"hub:commands","t":"cmd:changed-my-mind","k":"contains","h":1},{"s":"cmd:claude-md","t":"skill:second-brain-project","k":"follows"},{"s":"hub:commands","t":"cmd:claude-md","k":"contains","h":1},{"s":"cmd:clusters","t":"skill:second-brain-graph","k":"follows"},{"s":"hub:commands","t":"cmd:clusters","k":"contains","h":1},{"s":"cmd:commit","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:commit","k":"contains","h":1},{"s":"cmd:commitments","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:commitments","k":"contains","h":1},{"s":"cmd:compare","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:compare","k":"contains","h":1},{"s":"cmd:connect","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:connect","k":"contains","h":1},{"s":"cmd:contradictions","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:contradictions","k":"contains","h":1},{"s":"cmd:contradicts","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:contradicts","k":"contains","h":1},{"s":"cmd:decisions","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:decisions","k":"contains","h":1},{"s":"cmd:dedupe","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:dedupe","k":"contains","h":1},{"s":"cmd:doctor","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:doctor","k":"contains","h":1},{"s":"cmd:draft","t":"skill:second-brain-write","k":"follows"},{"s":"hub:commands","t":"cmd:draft","k":"contains","h":1},{"s":"cmd:dry-run","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:dry-run","k":"contains","h":1},{"s":"cmd:explain","t":"skill:second-brain-quiz","k":"follows"},{"s":"hub:commands","t":"cmd:explain","k":"contains","h":1},{"s":"cmd:export","t":"skill:second-brain-publish","k":"follows"},{"s":"hub:commands","t":"cmd:export","k":"contains","h":1},{"s":"cmd:gaps","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:gaps","k":"contains","h":1},{"s":"cmd:graph-export","t":"skill:second-brain-graph","k":"follows"},{"s":"hub:commands","t":"cmd:graph-export","k":"contains","h":1},{"s":"cmd:graph","t":"skill:second-brain-graph","k":"follows"},{"s":"hub:commands","t":"cmd:graph","k":"contains","h":1},{"s":"cmd:handoff","t":"skill:second-brain-report","k":"follows"},{"s":"hub:commands","t":"cmd:handoff","k":"contains","h":1},{"s":"cmd:health","t":"skill:second-brain-graph","k":"follows"},{"s":"hub:commands","t":"cmd:health","k":"contains","h":1},{"s":"cmd:hubs","t":"skill:second-brain-graph","k":"follows"},{"s":"hub:commands","t":"cmd:hubs","k":"contains","h":1},{"s":"cmd:index","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:index","k":"contains","h":1},{"s":"cmd:ingest-chats","t":"skill:second-brain-chat-import","k":"follows"},{"s":"hub:commands","t":"cmd:ingest-chats","k":"contains","h":1},{"s":"cmd:ingest-highlights","t":"skill:second-brain-ingest","k":"follows"},{"s":"hub:commands","t":"cmd:ingest-highlights","k":"contains","h":1},{"s":"cmd:ingest-mine","t":"skill:second-brain-ingest","k":"follows"},{"s":"hub:commands","t":"cmd:ingest-mine","k":"contains","h":1},{"s":"cmd:ingest-newsletter","t":"skill:second-brain-ingest","k":"follows"},{"s":"hub:commands","t":"cmd:ingest-newsletter","k":"contains","h":1},{"s":"cmd:ingest-paper","t":"skill:second-brain-ingest","k":"follows"},{"s":"hub:commands","t":"cmd:ingest-paper","k":"contains","h":1},{"s":"cmd:ingest-pdf","t":"skill:second-brain-ingest","k":"follows"},{"s":"hub:commands","t":"cmd:ingest-pdf","k":"contains","h":1},{"s":"cmd:ingest-url","t":"skill:second-brain-ingest","k":"follows"},{"s":"hub:commands","t":"cmd:ingest-url","k":"contains","h":1},{"s":"cmd:ingest-voice","t":"skill:second-brain-transcript","k":"follows"},{"s":"hub:commands","t":"cmd:ingest-voice","k":"contains","h":1},{"s":"cmd:ingest-youtube","t":"skill:second-brain-transcript","k":"follows"},{"s":"hub:commands","t":"cmd:ingest-youtube","k":"contains","h":1},{"s":"cmd:ingest","t":"skill:second-brain-ingest","k":"follows"},{"s":"hub:commands","t":"cmd:ingest","k":"contains","h":1},{"s":"cmd:init","t":"skill:second-brain-project","k":"follows"},{"s":"hub:commands","t":"cmd:init","k":"contains","h":1},{"s":"hub:commands","t":"cmd:install","k":"contains"},{"s":"cmd:know","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:know","k":"contains","h":1},{"s":"cmd:link","t":"skill:second-brain-ingest","k":"follows"},{"s":"hub:commands","t":"cmd:link","k":"contains","h":1},{"s":"cmd:lint","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:lint","k":"contains","h":1},{"s":"cmd:merge","t":"skill:second-brain-merge","k":"follows"},{"s":"hub:commands","t":"cmd:merge","k":"contains","h":1},{"s":"cmd:metrics","t":"skill:second-brain-metrics","k":"follows"},{"s":"hub:commands","t":"cmd:metrics","k":"contains","h":1},{"s":"cmd:monthly","t":"skill:second-brain-metrics","k":"follows"},{"s":"hub:commands","t":"cmd:monthly","k":"contains","h":1},{"s":"cmd:orphans","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:orphans","k":"contains","h":1},{"s":"cmd:outline","t":"skill:second-brain-write","k":"follows"},{"s":"hub:commands","t":"cmd:outline","k":"contains","h":1},{"s":"cmd:privacy","t":"skill:second-brain-privacy","k":"follows"},{"s":"hub:commands","t":"cmd:privacy","k":"contains","h":1},{"s":"cmd:project-status","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:project-status","k":"contains","h":1},{"s":"cmd:project","t":"skill:second-brain-project","k":"follows"},{"s":"hub:commands","t":"cmd:project","k":"contains","h":1},{"s":"cmd:prune","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:prune","k":"contains","h":1},{"s":"cmd:publish","t":"skill:second-brain-publish","k":"follows"},{"s":"hub:commands","t":"cmd:publish","k":"contains","h":1},{"s":"cmd:quiz","t":"skill:second-brain-quiz","k":"follows"},{"s":"hub:commands","t":"cmd:quiz","k":"contains","h":1},{"s":"cmd:rename","t":"skill:second-brain-rename","k":"follows"},{"s":"hub:commands","t":"cmd:rename","k":"contains","h":1},{"s":"cmd:report","t":"skill:second-brain-report","k":"follows"},{"s":"hub:commands","t":"cmd:report","k":"contains","h":1},{"s":"cmd:retype","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:retype","k":"contains","h":1},{"s":"cmd:review","t":"skill:second-brain-review","k":"follows"},{"s":"hub:commands","t":"cmd:review","k":"contains","h":1},{"s":"cmd:rollback","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:rollback","k":"contains","h":1},{"s":"cmd:schedule","t":"skill:second-brain-review","k":"follows"},{"s":"hub:commands","t":"cmd:schedule","k":"contains","h":1},{"s":"cmd:schema","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:schema","k":"contains","h":1},{"s":"cmd:scope","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:scope","k":"contains","h":1},{"s":"cmd:secrets","t":"skill:second-brain-privacy","k":"follows"},{"s":"hub:commands","t":"cmd:secrets","k":"contains","h":1},{"s":"cmd:sources","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:sources","k":"contains","h":1},{"s":"cmd:split","t":"skill:second-brain-ingest","k":"follows"},{"s":"hub:commands","t":"cmd:split","k":"contains","h":1},{"s":"cmd:stale","t":"skill:second-brain-graph","k":"follows"},{"s":"hub:commands","t":"cmd:stale","k":"contains","h":1},{"s":"cmd:tags","t":"skill:second-brain-lint","k":"follows"},{"s":"hub:commands","t":"cmd:tags","k":"contains","h":1},{"s":"cmd:timeline","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:timeline","k":"contains","h":1},{"s":"cmd:trace","t":"skill:second-brain-query","k":"follows"},{"s":"hub:commands","t":"cmd:trace","k":"contains","h":1},{"s":"cmd:typed-links","t":"skill:second-brain-ingest","k":"follows"},{"s":"hub:commands","t":"cmd:typed-links","k":"contains","h":1},{"s":"cmd:weekly","t":"skill:second-brain-review","k":"follows"},{"s":"hub:commands","t":"cmd:weekly","k":"contains","h":1},{"s":"hub:agents","t":"agent:curator","k":"contains"},{"s":"hub:agents","t":"agent:graph-analyst","k":"contains"},{"s":"hub:agents","t":"agent:ingestor","k":"contains"},{"s":"hub:agents","t":"agent:linker","k":"contains"},{"s":"hub:agents","t":"agent:researcher","k":"contains"},{"s":"hub:agents","t":"agent:reviewer","k":"contains"},{"s":"hub:scripts","t":"script:chat_export_to_md.py","k":"contains"},{"s":"hub:scripts","t":"script:graph_export.py","k":"contains"},{"s":"hub:scripts","t":"script:link_check.py","k":"contains"},{"s":"hub:scripts","t":"script:vault_stats.py","k":"contains"},{"s":"hub:plugin","t":"plug:context-audit","k":"contains"},{"s":"hub:plugin","t":"plug:evals-bootstrap","k":"contains"},{"s":"hub:plugin","t":"plug:gate-check","k":"contains"},{"s":"hub:plugin","t":"plug:goal-test","k":"contains"},{"s":"hub:plugin","t":"plug:harness-audit","k":"contains"},{"s":"hub:plugin","t":"plug:loop-critic","k":"contains"},{"s":"col:resources","t":"rcat:papers","k":"contains"},{"s":"col:resources","t":"rcat:tools","k":"contains"},{"s":"col:resources","t":"rcat:plugins","k":"contains"},{"s":"col:resources","t":"rcat:repositories","k":"contains"},{"s":"col:resources","t":"rcat:skills","k":"contains"},{"s":"col:resources","t":"rcat:reading","k":"contains"},{"s":"rcat:papers","t":"res:0","k":"contains"},{"s":"rcat:papers","t":"res:1","k":"contains"},{"s":"rcat:papers","t":"res:2","k":"contains"},{"s":"rcat:papers","t":"res:3","k":"contains"},{"s":"rcat:papers","t":"res:4","k":"contains"},{"s":"rcat:papers","t":"res:5","k":"contains"},{"s":"rcat:papers","t":"res:6","k":"contains"},{"s":"rcat:papers","t":"res:7","k":"contains"},{"s":"rcat:papers","t":"res:8","k":"contains"},{"s":"rcat:papers","t":"res:9","k":"contains"},{"s":"rcat:papers","t":"res:10","k":"contains"},{"s":"rcat:tools","t":"res:11","k":"contains"},{"s":"rcat:tools","t":"res:12","k":"contains"},{"s":"rcat:tools","t":"res:13","k":"contains"},{"s":"rcat:tools","t":"res:14","k":"contains"},{"s":"rcat:tools","t":"res:15","k":"contains"},{"s":"rcat:tools","t":"res:16","k":"contains"},{"s":"rcat:tools","t":"res:17","k":"contains"},{"s":"rcat:tools","t":"res:18","k":"contains"},{"s":"rcat:tools","t":"res:19","k":"contains"},{"s":"rcat:tools","t":"res:20","k":"contains"},{"s":"rcat:tools","t":"res:21","k":"contains"},{"s":"rcat:tools","t":"res:22","k":"contains"},{"s":"rcat:tools","t":"res:23","k":"contains"},{"s":"rcat:tools","t":"res:24","k":"contains"},{"s":"rcat:tools","t":"res:25","k":"contains"},{"s":"rcat:tools","t":"res:26","k":"contains"},{"s":"rcat:tools","t":"res:27","k":"contains"},{"s":"rcat:tools","t":"res:28","k":"contains"},{"s":"rcat:tools","t":"res:29","k":"contains"},{"s":"rcat:tools","t":"res:30","k":"contains"},{"s":"rcat:tools","t":"res:31","k":"contains"},{"s":"rcat:tools","t":"res:32","k":"contains"},{"s":"rcat:tools","t":"res:33","k":"contains"},{"s":"rcat:tools","t":"res:34","k":"contains"},{"s":"rcat:tools","t":"res:35","k":"contains"},{"s":"rcat:tools","t":"res:36","k":"contains"},{"s":"rcat:tools","t":"res:37","k":"contains"},{"s":"rcat:tools","t":"res:38","k":"contains"},{"s":"rcat:plugins","t":"res:39","k":"contains"},{"s":"rcat:plugins","t":"res:40","k":"contains"},{"s":"rcat:plugins","t":"res:41","k":"contains"},{"s":"rcat:plugins","t":"res:42","k":"contains"},{"s":"rcat:plugins","t":"res:43","k":"contains"},{"s":"rcat:plugins","t":"res:44","k":"contains"},{"s":"rcat:plugins","t":"res:45","k":"contains"},{"s":"rcat:plugins","t":"res:46","k":"contains"},{"s":"rcat:plugins","t":"res:47","k":"contains"},{"s":"rcat:plugins","t":"res:48","k":"contains"},{"s":"rcat:plugins","t":"res:49","k":"contains"},{"s":"rcat:plugins","t":"res:50","k":"contains"},{"s":"rcat:plugins","t":"res:51","k":"contains"},{"s":"rcat:plugins","t":"res:52","k":"contains"},{"s":"rcat:plugins","t":"res:53","k":"contains"},{"s":"rcat:plugins","t":"res:54","k":"contains"},{"s":"rcat:plugins","t":"res:55","k":"contains"},{"s":"rcat:plugins","t":"res:56","k":"contains"},{"s":"rcat:plugins","t":"res:57","k":"contains"},{"s":"rcat:plugins","t":"res:58","k":"contains"},{"s":"rcat:plugins","t":"res:59","k":"contains"},{"s":"rcat:plugins","t":"res:60","k":"contains"},{"s":"rcat:plugins","t":"res:61","k":"contains"},{"s":"rcat:plugins","t":"res:62","k":"contains"},{"s":"rcat:plugins","t":"res:63","k":"contains"},{"s":"rcat:plugins","t":"res:64","k":"contains"},{"s":"rcat:repositories","t":"res:65","k":"contains"},{"s":"rcat:repositories","t":"res:66","k":"contains"},{"s":"rcat:repositories","t":"res:67","k":"contains"},{"s":"rcat:repositories","t":"res:68","k":"contains"},{"s":"rcat:repositories","t":"res:69","k":"contains"},{"s":"rcat:repositories","t":"res:70","k":"contains"},{"s":"rcat:repositories","t":"res:71","k":"contains"},{"s":"rcat:repositories","t":"res:72","k":"contains"},{"s":"rcat:repositories","t":"res:73","k":"contains"},{"s":"rcat:repositories","t":"res:74","k":"contains"},{"s":"rcat:repositories","t":"res:75","k":"contains"},{"s":"rcat:repositories","t":"res:76","k":"contains"},{"s":"rcat:repositories","t":"res:77","k":"contains"},{"s":"rcat:repositories","t":"res:78","k":"contains"},{"s":"rcat:repositories","t":"res:79","k":"contains"},{"s":"rcat:repositories","t":"res:80","k":"contains"},{"s":"rcat:repositories","t":"res:81","k":"contains"},{"s":"rcat:repositories","t":"res:82","k":"contains"},{"s":"rcat:repositories","t":"res:83","k":"contains"},{"s":"rcat:repositories","t":"res:84","k":"contains"},{"s":"rcat:repositories","t":"res:85","k":"contains"},{"s":"rcat:repositories","t":"res:86","k":"contains"},{"s":"rcat:repositories","t":"res:87","k":"contains"},{"s":"rcat:skills","t":"res:88","k":"contains"},{"s":"rcat:skills","t":"res:89","k":"contains"},{"s":"rcat:skills","t":"res:90","k":"contains"},{"s":"rcat:skills","t":"res:91","k":"contains"},{"s":"rcat:skills","t":"res:92","k":"contains"},{"s":"rcat:skills","t":"res:93","k":"contains"},{"s":"rcat:skills","t":"res:94","k":"contains"},{"s":"rcat:skills","t":"res:95","k":"contains"},{"s":"rcat:skills","t":"res:96","k":"contains"},{"s":"rcat:skills","t":"res:97","k":"contains"},{"s":"rcat:skills","t":"res:98","k":"contains"},{"s":"rcat:skills","t":"res:99","k":"contains"},{"s":"rcat:skills","t":"res:100","k":"contains"},{"s":"rcat:skills","t":"res:101","k":"contains"},{"s":"rcat:skills","t":"res:102","k":"contains"},{"s":"rcat:skills","t":"res:103","k":"contains"},{"s":"rcat:skills","t":"res:104","k":"contains"},{"s":"rcat:reading","t":"res:105","k":"contains"},{"s":"rcat:reading","t":"res:106","k":"contains"},{"s":"rcat:reading","t":"res:107","k":"contains"},{"s":"rcat:reading","t":"res:108","k":"contains"},{"s":"rcat:reading","t":"res:109","k":"contains"},{"s":"rcat:reading","t":"res:110","k":"contains"},{"s":"rcat:reading","t":"res:111","k":"contains"},{"s":"rcat:reading","t":"res:112","k":"contains"},{"s":"rcat:reading","t":"res:113","k":"contains"},{"s":"course-0-map/what-an-agent-is","t":"course-0-map/agents-or-workflows","k":"link"},{"s":"course-0-map/what-an-agent-is","t":"course-0-map/five-layers","k":"link"},{"s":"course-0-map/five-layers","t":"course-1-context/how-models-read","k":"link"},{"s":"course-0-map/five-layers","t":"course-2-loop/loops-vs-workflows","k":"link"},{"s":"course-0-map/five-layers","t":"track-loop/what-loop-engineering-is","k":"link"},{"s":"course-0-map/five-layers","t":"course-3-gate/cheap-decisions","k":"link"},{"s":"course-0-map/five-layers","t":"track-jev/system-one-models","k":"link"},{"s":"course-0-map/five-layers","t":"course-4-harness/the-office","k":"link"},{"s":"course-0-map/five-layers","t":"track-harness/what-a-harness-is","k":"link"},{"s":"course-0-map/five-layers","t":"course-5-evals/two-kinds-of-checks","k":"link"},{"s":"course-0-map/five-layers","t":"track-evals/why-evals","k":"link"},{"s":"course-0-map/agents-or-workflows","t":"course-0-map/five-layers","k":"link"},{"s":"course-0-map/agents-or-workflows","t":"course-1-context/how-models-read","k":"link"},{"s":"course-0-map/agents-or-workflows","t":"course-2-loop/loops-vs-workflows","k":"link"},{"s":"course-0-map/agents-or-workflows","t":"track-loop/what-loop-engineering-is","k":"link"},{"s":"course-0-map/agents-or-workflows","t":"course-3-gate/cheap-decisions","k":"link"},{"s":"course-0-map/agents-or-workflows","t":"track-jev/system-one-models","k":"link"},{"s":"course-0-map/agents-or-workflows","t":"course-4-harness/the-office","k":"link"},{"s":"course-0-map/agents-or-workflows","t":"track-harness/what-a-harness-is","k":"link"},{"s":"course-0-map/agents-or-workflows","t":"course-5-evals/two-kinds-of-checks","k":"link"},{"s":"course-0-map/agents-or-workflows","t":"track-evals/why-evals","k":"link"},{"s":"course-0-map/agents-or-workflows","t":"course-6-production/from-prototype","k":"link"},{"s":"course-1-context/how-models-read","t":"course-1-context/the-four-places","k":"link"},{"s":"course-1-context/how-models-read","t":"track-harness/context-engineering","k":"link"},{"s":"course-1-context/the-four-places","t":"course-1-context/context-practice","k":"link"},{"s":"course-1-context/the-four-places","t":"track-loop/context-hygiene","k":"link"},{"s":"course-1-context/context-practice","t":"track-harness/context-engineering","k":"link"},{"s":"course-1-context/context-practice","t":"course-2-loop/loops-vs-workflows","k":"link"},{"s":"course-2-loop/loops-vs-workflows","t":"course-2-loop/the-four-parts","k":"link"},{"s":"course-2-loop/loops-vs-workflows","t":"track-loop/what-loop-engineering-is","k":"link"},{"s":"course-2-loop/loops-vs-workflows","t":"course-2-loop/loop-practice","k":"link"},{"s":"course-2-loop/the-four-parts","t":"track-loop/build-goal-test","k":"link"},{"s":"course-2-loop/the-four-parts","t":"track-loop/stop-conditions","k":"link"},{"s":"course-2-loop/the-four-parts","t":"course-2-loop/loop-practice","k":"link"},{"s":"course-2-loop/loop-practice","t":"track-loop/build-goal-test","k":"link"},{"s":"course-2-loop/loop-practice","t":"track-loop/stop-conditions","k":"link"},{"s":"course-2-loop/loop-practice","t":"course-3-gate/cheap-decisions","k":"link"},{"s":"course-2-loop/loop-practice","t":"track-loop/what-loop-engineering-is","k":"link"},{"s":"course-3-gate/cheap-decisions","t":"track-jev/system-one-models","k":"link"},{"s":"course-3-gate/cheap-decisions","t":"course-3-gate/gates-in-practice","k":"link"},{"s":"course-3-gate/gates-in-practice","t":"track-jev/build-router","k":"link"},{"s":"course-3-gate/gates-in-practice","t":"course-3-gate/gate-practice","k":"link"},{"s":"course-3-gate/gates-in-practice","t":"track-jev/system-one-models","k":"link"},{"s":"course-3-gate/gate-practice","t":"track-jev/build-router","k":"link"},{"s":"course-3-gate/gate-practice","t":"course-4-harness/the-office","k":"link"},{"s":"course-3-gate/gate-practice","t":"track-jev/system-one-models","k":"link"},{"s":"course-4-harness/the-office","t":"track-harness/what-a-harness-is","k":"link"},{"s":"course-4-harness/the-office","t":"course-4-harness/the-four-rings","k":"link"},{"s":"course-4-harness/the-office","t":"course-4-harness/harness-practice","k":"link"},{"s":"course-4-harness/the-four-rings","t":"track-harness/tools-and-mcp","k":"link"},{"s":"course-4-harness/the-four-rings","t":"course-4-harness/harness-practice","k":"link"},{"s":"course-4-harness/harness-practice","t":"track-harness/build-the-loop","k":"link"},{"s":"course-4-harness/harness-practice","t":"track-harness/tools-and-mcp","k":"link"},{"s":"course-4-harness/harness-practice","t":"course-5-evals/two-kinds-of-checks","k":"link"},{"s":"course-4-harness/harness-practice","t":"track-harness/what-a-harness-is","k":"link"},{"s":"course-5-evals/two-kinds-of-checks","t":"track-evals/agent-evals","k":"link"},{"s":"course-5-evals/two-kinds-of-checks","t":"track-evals/build-suite","k":"link"},{"s":"course-5-evals/two-kinds-of-checks","t":"course-5-evals/judges-and-golden-sets","k":"link"},{"s":"course-5-evals/judges-and-golden-sets","t":"track-evals/llm-as-judge","k":"link"},{"s":"course-5-evals/judges-and-golden-sets","t":"course-5-evals/evals-practice","k":"link"},{"s":"course-5-evals/evals-practice","t":"track-evals/why-evals","k":"link"},{"s":"course-5-evals/evals-practice","t":"track-evals/build-suite","k":"link"},{"s":"course-5-evals/evals-practice","t":"track-evals/llm-as-judge","k":"link"},{"s":"course-5-evals/evals-practice","t":"course-6-production/from-prototype","k":"link"},{"s":"course-5-evals/evals-practice","t":"track-evals/build-traces","k":"link"},{"s":"course-5-evals/evals-practice","t":"track-evals/build-ci","k":"link"},{"s":"course-6-production/from-prototype","t":"course-2-loop/loops-vs-workflows","k":"link"},{"s":"course-6-production/from-prototype","t":"course-5-evals/two-kinds-of-checks","k":"link"},{"s":"course-6-production/from-prototype","t":"course-6-production/operating-agents","k":"link"},{"s":"course-6-production/operating-agents","t":"course-2-loop/loops-vs-workflows","k":"link"},{"s":"course-6-production/operating-agents","t":"course-4-harness/the-office","k":"link"},{"s":"course-6-production/operating-agents","t":"course-5-evals/two-kinds-of-checks","k":"link"},{"s":"course-6-production/day-one-plan","t":"course-1-context/how-models-read","k":"link"},{"s":"course-6-production/day-one-plan","t":"track-harness/context-engineering","k":"link"},{"s":"course-6-production/day-one-plan","t":"course-2-loop/loops-vs-workflows","k":"link"},{"s":"course-6-production/day-one-plan","t":"sec:track-loop","k":"link"},{"s":"course-6-production/day-one-plan","t":"course-3-gate/cheap-decisions","k":"link"},{"s":"course-6-production/day-one-plan","t":"sec:track-jev","k":"link"},{"s":"course-6-production/day-one-plan","t":"course-4-harness/the-office","k":"link"},{"s":"course-6-production/day-one-plan","t":"sec:track-harness","k":"link"},{"s":"course-6-production/day-one-plan","t":"course-5-evals/two-kinds-of-checks","k":"link"},{"s":"course-6-production/day-one-plan","t":"sec:track-evals","k":"link"},{"s":"track-graph/why-graphs","t":"05-graphs/graph-basics","k":"link"},{"s":"track-graph/why-graphs","t":"track-graph/graphrag","k":"link"},{"s":"track-graph/why-graphs","t":"track-graph/graph-stores","k":"link"},{"s":"track-graph/why-graphs","t":"track-graph/building-graphs-with-llms","k":"link"},{"s":"track-graph/graphrag","t":"track-graph/tools","k":"link"},{"s":"track-graph/graphrag","t":"track-graph/building-graphs-with-llms","k":"link"},{"s":"track-graph/building-graphs-with-llms","t":"track-graph/graph-stores","k":"link"},{"s":"track-graph/building-graphs-with-llms","t":"track-graph/tools","k":"link"},{"s":"track-graph/graph-stores","t":"05-graphs/typed-links","k":"link"},{"s":"track-graph/graph-stores","t":"track-graph/graphrag","k":"link"},{"s":"track-graph/graph-stores","t":"track-graph/tools","k":"link"},{"s":"track-graph/build-extract","t":"track-graph/build-query","k":"link"},{"s":"track-graph/build-extract","t":"track-graph/build-use","k":"link"},{"s":"track-graph/build-extract","t":"track-graph/why-graphs","k":"link"},{"s":"track-graph/build-extract","t":"track-graph/tools","k":"link"},{"s":"track-graph/build-extract","t":"track-graph/building-graphs-with-llms","k":"link"},{"s":"track-graph/build-query","t":"track-graph/graph-stores","k":"link"},{"s":"track-graph/build-query","t":"track-graph/graphrag","k":"link"},{"s":"track-graph/build-query","t":"track-graph/why-graphs","k":"link"},{"s":"track-graph/build-query","t":"track-graph/build-use","k":"link"},{"s":"track-graph/build-use","t":"track-graph/why-graphs","k":"link"},{"s":"track-graph/build-use","t":"track-graph/graph-stores","k":"link"},{"s":"track-graph/build-use","t":"track-graph/tools","k":"link"},{"s":"track-graph/resources","t":"track-graph/graphrag","k":"link"},{"s":"track-graph/resources","t":"track-graph/tools","k":"link"},{"s":"track-graph/resources","t":"track-graph/why-graphs","k":"link"},{"s":"track-graph/resources","t":"track-graph/graph-stores","k":"link"},{"s":"track-jev/system-one-models","t":"track-jev/what-jev-is-good-for","k":"link"},{"s":"track-jev/system-one-models","t":"track-jev/getting-started","k":"link"},{"s":"track-jev/system-one-models","t":"track-jev/resources","k":"link"},{"s":"track-jev/what-jev-is-good-for","t":"track-jev/jev-in-an-agent-stack","k":"link"},{"s":"track-jev/what-jev-is-good-for","t":"track-jev/getting-started","k":"link"},{"s":"track-jev/jev-in-an-agent-stack","t":"track-jev/system-one-models","k":"link"},{"s":"track-jev/jev-in-an-agent-stack","t":"track-jev/getting-started","k":"link"},{"s":"track-jev/getting-started","t":"track-jev/resources","k":"link"},{"s":"track-jev/build-decision-endpoint","t":"track-jev/getting-started","k":"link"},{"s":"track-jev/build-decision-endpoint","t":"track-jev/system-one-models","k":"link"},{"s":"track-jev/build-decision-endpoint","t":"track-jev/build-router","k":"link"},{"s":"track-jev/build-router","t":"track-jev/jev-in-an-agent-stack","k":"link"},{"s":"track-jev/build-router","t":"track-jev/build-swap-in-jev","k":"link"},{"s":"track-jev/build-swap-in-jev","t":"track-jev/build-decision-endpoint","k":"link"},{"s":"track-jev/build-swap-in-jev","t":"track-jev/getting-started","k":"link"},{"s":"track-jev/resources","t":"track-jev/jev-in-an-agent-stack","k":"link"},{"s":"track-jev/resources","t":"track-jev/what-jev-is-good-for","k":"link"},{"s":"track-harness/what-a-harness-is","t":"track-harness/tools-and-mcp","k":"link"},{"s":"track-harness/what-a-harness-is","t":"track-harness/context-engineering","k":"link"},{"s":"track-harness/what-a-harness-is","t":"track-harness/claude-code-as-harness","k":"link"},{"s":"track-harness/what-a-harness-is","t":"track-harness/harness-landscape","k":"link"},{"s":"track-harness/claude-code-as-harness","t":"track-harness/context-engineering","k":"link"},{"s":"track-harness/claude-code-as-harness","t":"track-harness/tools-and-mcp","k":"link"},{"s":"track-harness/claude-code-as-harness","t":"track-harness/harness-landscape","k":"link"},{"s":"track-harness/context-engineering","t":"track-harness/tools-and-mcp","k":"link"},{"s":"track-harness/tools-and-mcp","t":"track-harness/harness-landscape","k":"link"},{"s":"track-harness/harness-landscape","t":"track-harness/context-engineering","k":"link"},{"s":"track-harness/build-the-loop","t":"track-harness/what-a-harness-is","k":"link"},{"s":"track-harness/build-the-loop","t":"track-harness/build-guardrails","k":"link"},{"s":"track-harness/build-the-loop","t":"track-harness/tools-and-mcp","k":"link"},{"s":"track-harness/build-the-loop","t":"track-harness/resources","k":"link"},{"s":"track-harness/build-guardrails","t":"track-harness/what-a-harness-is","k":"link"},{"s":"track-harness/build-guardrails","t":"track-harness/context-engineering","k":"link"},{"s":"track-harness/build-guardrails","t":"track-harness/build-graduate","k":"link"},{"s":"track-harness/build-graduate","t":"track-harness/harness-landscape","k":"link"},{"s":"track-harness/build-graduate","t":"track-harness/context-engineering","k":"link"},{"s":"track-harness/build-graduate","t":"track-harness/claude-code-as-harness","k":"link"},{"s":"track-harness/resources","t":"track-harness/what-a-harness-is","k":"link"},{"s":"track-harness/resources","t":"track-harness/context-engineering","k":"link"},{"s":"track-harness/resources","t":"track-harness/tools-and-mcp","k":"link"},{"s":"track-harness/resources","t":"track-harness/harness-landscape","k":"link"},{"s":"track-loop/what-loop-engineering-is","t":"track-loop/context-hygiene","k":"link"},{"s":"track-loop/what-loop-engineering-is","t":"track-loop/stop-conditions","k":"link"},{"s":"track-loop/what-loop-engineering-is","t":"track-loop/critics-and-verification","k":"link"},{"s":"track-loop/what-loop-engineering-is","t":"track-harness/what-a-harness-is","k":"link"},{"s":"track-loop/stop-conditions","t":"track-loop/build-goal-test","k":"link"},{"s":"track-loop/stop-conditions","t":"track-loop/critics-and-verification","k":"link"},{"s":"track-loop/stop-conditions","t":"track-loop/build-overnight","k":"link"},{"s":"track-loop/stop-conditions","t":"track-loop/patterns","k":"link"},{"s":"track-loop/critics-and-verification","t":"track-evals/llm-as-judge","k":"link"},{"s":"track-loop/critics-and-verification","t":"track-loop/build-critic","k":"link"},{"s":"track-loop/context-hygiene","t":"track-harness/context-engineering","k":"link"},{"s":"track-loop/context-hygiene","t":"track-loop/build-overnight","k":"link"},{"s":"track-loop/context-hygiene","t":"track-loop/build-goal-test","k":"link"},{"s":"track-loop/patterns","t":"track-loop/critics-and-verification","k":"link"},{"s":"track-loop/patterns","t":"track-loop/build-critic","k":"link"},{"s":"track-loop/patterns","t":"track-loop/build-goal-test","k":"link"},{"s":"track-loop/build-goal-test","t":"track-harness/claude-code-as-harness","k":"link"},{"s":"track-loop/build-goal-test","t":"track-loop/critics-and-verification","k":"link"},{"s":"track-loop/build-goal-test","t":"track-loop/build-critic","k":"link"},{"s":"track-loop/build-critic","t":"track-loop/build-overnight","k":"link"},{"s":"track-loop/build-overnight","t":"track-loop/build-goal-test","k":"link"},{"s":"track-loop/resources","t":"track-loop/context-hygiene","k":"link"},{"s":"track-loop/resources","t":"track-harness/resources","k":"link"},{"s":"track-loop/resources","t":"track-loop/what-loop-engineering-is","k":"link"},{"s":"track-loop/resources","t":"track-loop/build-goal-test","k":"link"},{"s":"track-evals/why-evals","t":"track-evals/llm-as-judge","k":"link"},{"s":"track-evals/why-evals","t":"track-evals/designing-evals","k":"link"},{"s":"track-evals/why-evals","t":"track-evals/agent-evals","k":"link"},{"s":"track-evals/designing-evals","t":"track-evals/llm-as-judge","k":"link"},{"s":"track-evals/designing-evals","t":"track-evals/agent-evals","k":"link"},{"s":"track-evals/llm-as-judge","t":"track-evals/agent-evals","k":"link"},{"s":"track-evals/agent-evals","t":"track-evals/tooling","k":"link"},{"s":"track-evals/tooling","t":"track-evals/why-evals","k":"link"},{"s":"track-evals/tooling","t":"track-evals/llm-as-judge","k":"link"},{"s":"track-evals/build-traces","t":"track-evals/build-suite","k":"link"},{"s":"track-evals/build-traces","t":"track-evals/why-evals","k":"link"},{"s":"track-evals/build-traces","t":"track-evals/designing-evals","k":"link"},{"s":"track-evals/build-suite","t":"track-evals/llm-as-judge","k":"link"},{"s":"track-evals/build-suite","t":"track-evals/tooling","k":"link"},{"s":"track-evals/build-ci","t":"track-evals/build-suite","k":"link"},{"s":"track-evals/build-ci","t":"track-evals/agent-evals","k":"link"},{"s":"track-evals/build-ci","t":"track-evals/designing-evals","k":"link"},{"s":"track-evals/build-ci","t":"track-evals/why-evals","k":"link"},{"s":"track-evals/resources","t":"track-evals/why-evals","k":"link"},{"s":"track-evals/resources","t":"track-evals/agent-evals","k":"link"},{"s":"track-evals/resources","t":"track-evals/designing-evals","k":"link"},{"s":"track-evals/resources","t":"track-evals/llm-as-judge","k":"link"},{"s":"01-concepts/what-is-a-second-brain","t":"01-concepts/what-good-looks-like","k":"link"},{"s":"01-concepts/what-is-a-second-brain","t":"01-concepts/the-save-for-later-paradox","k":"link"},{"s":"01-concepts/the-save-for-later-paradox","t":"01-concepts/llm-wiki-pattern","k":"link"},{"s":"01-concepts/llm-wiki-pattern","t":"sec:04-structuring","k":"link"},{"s":"01-concepts/llm-wiki-pattern","t":"01-concepts/two-layers","k":"link"},{"s":"01-concepts/llm-wiki-pattern","t":"sec:06-agents","k":"link"},{"s":"01-concepts/two-layers","t":"02-setup/vault-structure","k":"link"},{"s":"01-concepts/why-markdown-and-plain-text","t":"05-graphs/exporting-your-graph","k":"link"},{"s":"01-concepts/why-markdown-and-plain-text","t":"01-concepts/pkm-lineage","k":"link"},{"s":"01-concepts/pkm-lineage","t":"sec:04-structuring","k":"link"},{"s":"01-concepts/pkm-lineage","t":"01-concepts/two-layers","k":"link"},{"s":"01-concepts/pkm-lineage","t":"01-concepts/wiki-vs-rag","k":"link"},{"s":"01-concepts/wiki-vs-rag","t":"07-retrieval/rag-on-top","k":"link"},{"s":"01-concepts/wiki-vs-rag","t":"05-graphs/graph-vs-vectors","k":"link"},{"s":"01-concepts/wiki-vs-rag","t":"01-concepts/what-good-looks-like","k":"link"},{"s":"01-concepts/what-good-looks-like","t":"05-graphs/metrics","k":"link"},{"s":"01-concepts/what-good-looks-like","t":"sec:02-setup","k":"link"},{"s":"02-setup/obsidian-install-and-vault","t":"02-setup/git-and-sync","k":"link"},{"s":"02-setup/obsidian-install-and-vault","t":"02-setup/claude-code-setup","k":"link"},{"s":"02-setup/claude-code-setup","t":"02-setup/mcp-obsidian","k":"link"},{"s":"02-setup/claude-code-setup","t":"02-setup/claude-md","k":"link"},{"s":"02-setup/claude-code-setup","t":"06-agents/safety-and-guardrails","k":"link"},{"s":"02-setup/mcp-obsidian","t":"02-setup/claude-md","k":"link"},{"s":"02-setup/claude-md","t":"02-setup/project-scoping","k":"link"},{"s":"02-setup/claude-md","t":"02-setup/vault-structure","k":"link"},{"s":"02-setup/vault-structure","t":"02-setup/project-scoping","k":"link"},{"s":"02-setup/project-scoping","t":"02-setup/live-data","k":"link"},{"s":"02-setup/live-data","t":"06-agents/safety-and-guardrails","k":"link"},{"s":"02-setup/live-data","t":"02-setup/obsidian-plugins","k":"link"},{"s":"02-setup/obsidian-plugins","t":"02-setup/mcp-obsidian","k":"link"},{"s":"02-setup/obsidian-plugins","t":"09-maintenance/versioning-with-git","k":"link"},{"s":"02-setup/obsidian-plugins","t":"09-maintenance/lint-and-health","k":"link"},{"s":"02-setup/obsidian-plugins","t":"02-setup/git-and-sync","k":"link"},{"s":"02-setup/git-and-sync","t":"sec:03-ingestion","k":"link"},{"s":"02-setup/git-and-sync","t":"sec:04-structuring","k":"link"},{"s":"03-ingestion/web-clipper","t":"03-ingestion/youtube-transcripts","k":"link"},{"s":"03-ingestion/youtube-transcripts","t":"03-ingestion/pdfs-and-books","k":"link"},{"s":"03-ingestion/pdfs-and-books","t":"03-ingestion/chat-exports","k":"link"},{"s":"03-ingestion/chat-exports","t":"09-maintenance/privacy-and-secrets","k":"link"},{"s":"03-ingestion/chat-exports","t":"03-ingestion/voice-and-meetings","k":"link"},{"s":"03-ingestion/voice-and-meetings","t":"01-concepts/two-layers","k":"link"},{"s":"03-ingestion/voice-and-meetings","t":"03-ingestion/newsletters-and-email","k":"link"},{"s":"03-ingestion/newsletters-and-email","t":"03-ingestion/bulk-backfill","k":"link"},{"s":"03-ingestion/bulk-backfill","t":"sec:04-structuring","k":"link"},{"s":"04-structuring/atomic-pages","t":"04-structuring/page-types","k":"link"},{"s":"04-structuring/page-types","t":"04-structuring/linking-rules","k":"link"},{"s":"04-structuring/linking-rules","t":"04-structuring/frontmatter-schema","k":"link"},{"s":"04-structuring/frontmatter-schema","t":"04-structuring/naming-and-aliases","k":"link"},{"s":"04-structuring/naming-and-aliases","t":"04-structuring/dedupe-and-merge","k":"link"},{"s":"04-structuring/dedupe-and-merge","t":"04-structuring/contradictions-and-supersession","k":"link"},{"s":"04-structuring/contradictions-and-supersession","t":"04-structuring/index-and-log","k":"link"},{"s":"04-structuring/index-and-log","t":"sec:05-graphs","k":"link"},{"s":"05-graphs/graph-basics","t":"05-graphs/obsidian-graph-view","k":"link"},{"s":"05-graphs/obsidian-graph-view","t":"05-graphs/typed-links","k":"link"},{"s":"05-graphs/typed-links","t":"05-graphs/graph-vs-vectors","k":"link"},{"s":"05-graphs/graph-vs-vectors","t":"07-retrieval/rag-on-top","k":"link"},{"s":"05-graphs/graph-vs-vectors","t":"05-graphs/graphrag","k":"link"},{"s":"05-graphs/graphrag","t":"05-graphs/exporting-your-graph","k":"link"},{"s":"05-graphs/exporting-your-graph","t":"05-graphs/metrics","k":"link"},{"s":"05-graphs/metrics","t":"sec:06-agents","k":"link"},{"s":"06-agents/agent-roles","t":"06-agents/scheduled-maintenance","k":"link"},{"s":"06-agents/scheduled-maintenance","t":"06-agents/safety-and-guardrails","k":"link"},{"s":"06-agents/scheduled-maintenance","t":"05-graphs/metrics","k":"link"},{"s":"06-agents/scheduled-maintenance","t":"06-agents/subagents","k":"link"},{"s":"06-agents/subagents","t":"06-agents/hooks","k":"link"},{"s":"06-agents/hooks","t":"06-agents/skills-and-commands","k":"link"},{"s":"06-agents/skills-and-commands","t":"06-agents/safety-and-guardrails","k":"link"},{"s":"06-agents/safety-and-guardrails","t":"09-maintenance/privacy-and-secrets","k":"link"},{"s":"06-agents/safety-and-guardrails","t":"sec:07-retrieval","k":"link"},{"s":"07-retrieval/asking-questions","t":"04-structuring/index-and-log","k":"link"},{"s":"07-retrieval/asking-questions","t":"07-retrieval/query-patterns","k":"link"},{"s":"07-retrieval/query-patterns","t":"04-structuring/contradictions-and-supersession","k":"link"},{"s":"07-retrieval/query-patterns","t":"03-ingestion/chat-exports","k":"link"},{"s":"07-retrieval/query-patterns","t":"07-retrieval/search-tools","k":"link"},{"s":"07-retrieval/search-tools","t":"07-retrieval/context-budget","k":"link"},{"s":"07-retrieval/search-tools","t":"07-retrieval/rag-on-top","k":"link"},{"s":"07-retrieval/rag-on-top","t":"07-retrieval/context-budget","k":"link"},{"s":"07-retrieval/context-budget","t":"06-agents/subagents","k":"link"},{"s":"07-retrieval/context-budget","t":"sec:08-outputs","k":"link"},{"s":"08-outputs/writing-from-the-vault","t":"08-outputs/research-reports","k":"link"},{"s":"08-outputs/research-reports","t":"04-structuring/contradictions-and-supersession","k":"link"},{"s":"08-outputs/research-reports","t":"08-outputs/publishing-and-export","k":"link"},{"s":"08-outputs/publishing-and-export","t":"08-outputs/teaching-yourself","k":"link"},{"s":"08-outputs/teaching-yourself","t":"sec:09-maintenance","k":"link"},{"s":"09-maintenance/lint-and-health","t":"05-graphs/metrics","k":"link"},{"s":"09-maintenance/lint-and-health","t":"09-maintenance/review-cadence","k":"link"},{"s":"09-maintenance/review-cadence","t":"05-graphs/metrics","k":"link"},{"s":"09-maintenance/review-cadence","t":"04-structuring/contradictions-and-supersession","k":"link"},{"s":"09-maintenance/review-cadence","t":"09-maintenance/versioning-with-git","k":"link"},{"s":"09-maintenance/versioning-with-git","t":"06-agents/scheduled-maintenance","k":"link"},{"s":"09-maintenance/versioning-with-git","t":"09-maintenance/backups-and-portability","k":"link"},{"s":"09-maintenance/backups-and-portability","t":"09-maintenance/privacy-and-secrets","k":"link"},{"s":"09-maintenance/privacy-and-secrets","t":"08-outputs/publishing-and-export","k":"link"},{"s":"09-maintenance/privacy-and-secrets","t":"09-maintenance/scaling","k":"link"},{"s":"09-maintenance/scaling","t":"sec:10-troubleshooting","k":"link"},{"s":"10-troubleshooting/common-failures","t":"sec:03-ingestion","k":"link"},{"s":"10-troubleshooting/common-failures","t":"01-concepts/what-good-looks-like","k":"link"},{"s":"10-troubleshooting/common-failures","t":"04-structuring/linking-rules","k":"link"},{"s":"10-troubleshooting/common-failures","t":"10-troubleshooting/agent-writes-garbage","k":"link"},{"s":"10-troubleshooting/agent-writes-garbage","t":"03-ingestion/pdfs-and-books","k":"link"},{"s":"10-troubleshooting/agent-writes-garbage","t":"03-ingestion/youtube-transcripts","k":"link"},{"s":"10-troubleshooting/agent-writes-garbage","t":"04-structuring/page-types","k":"link"},{"s":"10-troubleshooting/agent-writes-garbage","t":"10-troubleshooting/broken-links","k":"link"},{"s":"10-troubleshooting/broken-links","t":"04-structuring/naming-and-aliases","k":"link"},{"s":"10-troubleshooting/broken-links","t":"10-troubleshooting/vault-too-big","k":"link"},{"s":"10-troubleshooting/vault-too-big","t":"05-graphs/metrics","k":"link"},{"s":"10-troubleshooting/vault-too-big","t":"09-maintenance/scaling","k":"link"},{"s":"10-troubleshooting/vault-too-big","t":"10-troubleshooting/faq","k":"link"},{"s":"10-troubleshooting/faq","t":"02-setup/mcp-obsidian","k":"link"},{"s":"10-troubleshooting/faq","t":"03-ingestion/bulk-backfill","k":"link"},{"s":"10-troubleshooting/faq","t":"sec:04-structuring","k":"link"},{"s":"10-troubleshooting/faq","t":"05-graphs/obsidian-graph-view","k":"link"},{"s":"course-1-context/how-models-read","t":"res:2","k":"cites"},{"s":"course-1-context/context-practice","t":"hub:plugin","k":"uses","w":3},{"s":"course-2-loop/loop-practice","t":"hub:plugin","k":"uses","w":3},{"s":"course-3-gate/gate-practice","t":"hub:plugin","k":"uses","w":3},{"s":"course-4-harness/harness-practice","t":"hub:plugin","k":"uses","w":3},{"s":"course-5-evals/evals-practice","t":"hub:plugin","k":"uses","w":3},{"s":"course-1-context/context-practice","t":"plug:context-audit","k":"uses"},{"s":"course-2-loop/loop-practice","t":"plug:goal-test","k":"uses","w":2},{"s":"course-2-loop/loop-practice","t":"plug:loop-critic","k":"uses"},{"s":"course-3-gate/gate-practice","t":"plug:gate-check","k":"uses"},{"s":"course-4-harness/harness-practice","t":"plug:harness-audit","k":"uses"},{"s":"course-5-evals/evals-practice","t":"plug:evals-bootstrap","k":"uses"},{"s":"track-graph/why-graphs","t":"res:65","k":"cites"},{"s":"track-graph/graphrag","t":"res:65","k":"cites"},{"s":"track-graph/tools","t":"res:65","k":"cites"},{"s":"track-graph/resources","t":"res:65","k":"cites"},{"s":"track-graph/why-graphs","t":"res:71","k":"cites"},{"s":"track-graph/building-graphs-with-llms","t":"res:71","k":"cites"},{"s":"track-graph/graph-stores","t":"res:71","k":"cites"},{"s":"track-graph/tools","t":"res:71","k":"cites"},{"s":"track-graph/resources","t":"res:71","k":"cites"},{"s":"track-graph/graphrag","t":"res:0","k":"cites"},{"s":"track-graph/resources","t":"res:0","k":"cites"},{"s":"track-graph/graphrag","t":"res:66","k":"cites"},{"s":"track-graph/tools","t":"res:66","k":"cites"},{"s":"track-graph/resources","t":"res:66","k":"cites"},{"s":"track-graph/graph-stores","t":"res:33","k":"cites"},{"s":"track-graph/graph-stores","t":"res:35","k":"cites"},{"s":"track-graph/tools","t":"res:72","k":"cites"},{"s":"track-graph/tools","t":"res:36","k":"cites"},{"s":"track-graph/build-extract","t":"script:graph_export.py","k":"uses","w":2},{"s":"track-graph/build-use","t":"script:graph_export.py","k":"uses","w":2},{"s":"05-graphs/exporting-your-graph","t":"script:graph_export.py","k":"uses","w":2},{"s":"01-concepts/why-markdown-and-plain-text","t":"script:graph_export.py","k":"uses"},{"s":"track-graph/resources","t":"res:1","k":"cites"},{"s":"track-evals/tooling","t":"res:100","k":"cites"},{"s":"01-concepts/llm-wiki-pattern","t":"rcat:skills","k":"uses","w":3},{"s":"02-setup/obsidian-install-and-vault","t":"res:11","k":"cites"},{"s":"02-setup/obsidian-install-and-vault","t":"hub:vault","k":"uses","w":3},{"s":"02-setup/claude-md","t":"hub:vault","k":"uses","w":3},{"s":"02-setup/obsidian-plugins","t":"hub:vault","k":"uses","w":3},{"s":"02-setup/claude-code-setup","t":"res:13","k":"cites"},{"s":"02-setup/claude-code-setup","t":"res:39","k":"cites"},{"s":"02-setup/claude-code-setup","t":"res:40","k":"cites"},{"s":"02-setup/mcp-obsidian","t":"res:42","k":"cites"},{"s":"02-setup/obsidian-plugins","t":"res:54","k":"cites"},{"s":"02-setup/obsidian-plugins","t":"res:56","k":"cites"},{"s":"10-troubleshooting/faq","t":"res:56","k":"cites"},{"s":"02-setup/obsidian-plugins","t":"res:58","k":"cites"},{"s":"09-maintenance/lint-and-health","t":"res:58","k":"cites"},{"s":"02-setup/obsidian-plugins","t":"res:55","k":"cites"},{"s":"02-setup/obsidian-plugins","t":"rcat:plugins","k":"uses","w":3},{"s":"03-ingestion/web-clipper","t":"res:12","k":"cites"},{"s":"03-ingestion/web-clipper","t":"cmd:ingest","k":"uses"},{"s":"06-agents/skills-and-commands","t":"cmd:ingest","k":"uses"},{"s":"03-ingestion/chat-exports","t":"hub:scripts","k":"uses","w":3},{"s":"03-ingestion/chat-exports","t":"script:chat_export_to_md.py","k":"uses"},{"s":"03-ingestion/bulk-backfill","t":"hub:commands","k":"uses","w":3},{"s":"06-agents/skills-and-commands","t":"hub:commands","k":"uses","w":3},{"s":"03-ingestion/bulk-backfill","t":"cmd:backfill","k":"uses"},{"s":"03-ingestion/bulk-backfill","t":"cmd:lint","k":"uses"},{"s":"09-maintenance/lint-and-health","t":"cmd:lint","k":"uses"},{"s":"04-structuring/linking-rules","t":"script:vault_stats.py","k":"uses"},{"s":"05-graphs/metrics","t":"script:vault_stats.py","k":"uses"},{"s":"04-structuring/naming-and-aliases","t":"script:link_check.py","k":"uses"},{"s":"07-retrieval/rag-on-top","t":"script:link_check.py","k":"uses"},{"s":"09-maintenance/lint-and-health","t":"script:link_check.py","k":"uses"},{"s":"09-maintenance/backups-and-portability","t":"script:link_check.py","k":"uses"},{"s":"10-troubleshooting/broken-links","t":"script:link_check.py","k":"uses"},{"s":"04-structuring/dedupe-and-merge","t":"skill:second-brain-lint","k":"uses"},{"s":"04-structuring/index-and-log","t":"hub:skills","k":"uses","w":3},{"s":"06-agents/agent-roles","t":"hub:skills","k":"uses","w":3},{"s":"06-agents/skills-and-commands","t":"hub:skills","k":"uses","w":3},{"s":"05-graphs/exporting-your-graph","t":"res:85","k":"cites"},{"s":"06-agents/agent-roles","t":"hub:agents","k":"uses","w":3},{"s":"06-agents/agent-roles","t":"agent:curator","k":"uses"},{"s":"06-agents/agent-roles","t":"agent:ingestor","k":"uses","w":5},{"s":"06-agents/agent-roles","t":"agent:linker","k":"uses","w":2},{"s":"06-agents/scheduled-maintenance","t":"agent:linker","k":"uses"},{"s":"06-agents/agent-roles","t":"agent:reviewer","k":"uses","w":3},{"s":"08-outputs/teaching-yourself","t":"res:61","k":"cites"},{"s":"09-maintenance/review-cadence","t":"cmd:review","k":"uses"},{"s":"col:glossary","t":"term:second-brain","k":"contains","h":1},{"s":"term:second-brain","t":"01-concepts/what-is-a-second-brain","k":"mentions","w":0},{"s":"term:second-brain","t":"track-graph/why-graphs","k":"mentions"},{"s":"term:second-brain","t":"track-harness/context-engineering","k":"mentions"},{"s":"term:second-brain","t":"track-loop/context-hygiene","k":"mentions"},{"s":"res:91","t":"term:second-brain","k":"related"},{"s":"res:92","t":"term:second-brain","k":"related"},{"s":"res:106","t":"term:second-brain","k":"related"},{"s":"col:glossary","t":"term:llm-wiki","k":"contains","h":1},{"s":"term:llm-wiki","t":"01-concepts/llm-wiki-pattern","k":"mentions","w":0},{"s":"term:llm-wiki","t":"01-concepts/the-save-for-later-paradox","k":"mentions"},{"s":"term:llm-wiki","t":"01-concepts/two-layers","k":"mentions"},{"s":"res:84","t":"term:llm-wiki","k":"related"},{"s":"res:94","t":"term:llm-wiki","k":"related"},{"s":"res:105","t":"term:llm-wiki","k":"related"},{"s":"col:glossary","t":"term:vault","k":"contains","h":1},{"s":"term:vault","t":"07-retrieval/asking-questions","k":"mentions","w":9},{"s":"term:vault","t":"02-setup/claude-code-setup","k":"mentions","w":8},{"s":"term:vault","t":"10-troubleshooting/common-failures","k":"mentions","w":7},{"s":"term:vault","t":"09-maintenance/privacy-and-secrets","k":"mentions","w":6},{"s":"term:vault","t":"02-setup/obsidian-install-and-vault","k":"mentions","w":5},{"s":"term:vault","t":"02-setup/vault-structure","k":"mentions","w":4},{"s":"term:vault","t":"08-outputs/writing-from-the-vault","k":"mentions","w":4},{"s":"term:vault","t":"05-graphs/graphrag","k":"mentions","w":3},{"s":"term:vault","t":"06-agents/safety-and-guardrails","k":"mentions","w":3},{"s":"term:vault","t":"07-retrieval/context-budget","k":"mentions","w":3},{"s":"col:glossary","t":"term:obsidian","k":"contains","h":1},{"s":"term:obsidian","t":"02-setup/mcp-obsidian","k":"mentions","w":7},{"s":"term:obsidian","t":"02-setup/obsidian-install-and-vault","k":"mentions","w":3},{"s":"term:obsidian","t":"02-setup/claude-code-setup","k":"mentions","w":3},{"s":"term:obsidian","t":"track-graph/tools","k":"mentions","w":2},{"s":"term:obsidian","t":"03-ingestion/web-clipper","k":"mentions","w":2},{"s":"term:obsidian","t":"07-retrieval/search-tools","k":"mentions"},{"s":"term:obsidian","t":"05-graphs/obsidian-graph-view","k":"mentions","w":0},{"s":"term:obsidian","t":"02-setup/git-and-sync","k":"mentions","w":5},{"s":"term:obsidian","t":"02-setup/obsidian-plugins","k":"mentions","w":3},{"s":"term:obsidian","t":"10-troubleshooting/faq","k":"mentions","w":3},{"s":"res:11","t":"term:obsidian","k":"related"},{"s":"res:12","t":"term:obsidian","k":"related"},{"s":"res:90","t":"term:obsidian","k":"related"},{"s":"res:93","t":"term:obsidian","k":"related"},{"s":"col:glossary","t":"term:claude-code","k":"contains","h":1},{"s":"term:claude-code","t":"02-setup/claude-code-setup","k":"mentions","w":6},{"s":"term:claude-code","t":"track-harness/claude-code-as-harness","k":"mentions","w":2},{"s":"term:claude-code","t":"track-evals/tooling","k":"mentions","w":2},{"s":"term:claude-code","t":"track-harness/build-graduate","k":"mentions","w":3},{"s":"term:claude-code","t":"course-4-harness/the-four-rings","k":"mentions","w":2},{"s":"term:claude-code","t":"track-graph/build-extract","k":"mentions","w":2},{"s":"term:claude-code","t":"track-harness/harness-landscape","k":"mentions","w":2},{"s":"term:claude-code","t":"10-troubleshooting/faq","k":"mentions","w":2},{"s":"res:13","t":"term:claude-code","k":"related"},{"s":"res:99","t":"term:claude-code","k":"related"},{"s":"res:113","t":"term:claude-code","k":"related"},{"s":"col:glossary","t":"term:claude-md","k":"contains","h":1},{"s":"term:claude-md","t":"02-setup/claude-md","k":"mentions","w":7},{"s":"term:claude-md","t":"track-harness/claude-code-as-harness","k":"mentions","w":2},{"s":"term:claude-md","t":"10-troubleshooting/agent-writes-garbage","k":"mentions","w":2},{"s":"term:claude-md","t":"02-setup/vault-structure","k":"mentions","w":4},{"s":"term:claude-md","t":"02-setup/claude-code-setup","k":"mentions","w":3},{"s":"term:claude-md","t":"02-setup/project-scoping","k":"mentions","w":3},{"s":"term:claude-md","t":"10-troubleshooting/common-failures","k":"mentions","w":2},{"s":"col:glossary","t":"term:mcp","k":"contains","h":1},{"s":"term:mcp","t":"02-setup/claude-code-setup","k":"mentions","w":5},{"s":"term:mcp","t":"track-harness/claude-code-as-harness","k":"mentions","w":4},{"s":"term:mcp","t":"track-harness/tools-and-mcp","k":"mentions","w":4},{"s":"term:mcp","t":"02-setup/mcp-obsidian","k":"mentions","w":3},{"s":"term:mcp","t":"course-4-harness/the-four-rings","k":"mentions","w":3},{"s":"term:mcp","t":"track-harness/resources","k":"mentions","w":3},{"s":"term:mcp","t":"10-troubleshooting/faq","k":"mentions","w":3},{"s":"term:mcp","t":"02-setup/live-data","k":"mentions","w":2},{"s":"term:mcp","t":"04-structuring/naming-and-aliases","k":"mentions","w":2},{"s":"res:42","t":"term:mcp","k":"related"},{"s":"col:glossary","t":"term:raw","k":"contains","h":1},{"s":"term:raw","t":"06-agents/scheduled-maintenance","k":"mentions","w":3},{"s":"term:raw","t":"02-setup/vault-structure","k":"mentions","w":2},{"s":"term:raw","t":"03-ingestion/web-clipper","k":"mentions","w":2},{"s":"term:raw","t":"03-ingestion/chat-exports","k":"mentions","w":2},{"s":"term:raw","t":"03-ingestion/newsletters-and-email","k":"mentions","w":2},{"s":"col:glossary","t":"term:two-layers","k":"contains","h":1},{"s":"term:two-layers","t":"01-concepts/two-layers","k":"mentions","w":5},{"s":"term:two-layers","t":"02-setup/vault-structure","k":"mentions","w":5},{"s":"term:two-layers","t":"01-concepts/llm-wiki-pattern","k":"mentions","w":3},{"s":"term:two-layers","t":"03-ingestion/voice-and-meetings","k":"mentions","w":2},{"s":"col:glossary","t":"term:ingest","k":"contains","h":1},{"s":"term:ingest","t":"03-ingestion/web-clipper","k":"mentions","w":7},{"s":"term:ingest","t":"03-ingestion/chat-exports","k":"mentions","w":5},{"s":"term:ingest","t":"10-troubleshooting/broken-links","k":"mentions","w":5},{"s":"term:ingest","t":"03-ingestion/youtube-transcripts","k":"mentions","w":4},{"s":"term:ingest","t":"10-troubleshooting/common-failures","k":"mentions","w":4},{"s":"term:ingest","t":"03-ingestion/newsletters-and-email","k":"mentions","w":6},{"s":"term:ingest","t":"06-agents/scheduled-maintenance","k":"mentions","w":5},{"s":"term:ingest","t":"01-concepts/wiki-vs-rag","k":"mentions","w":4},{"s":"term:ingest","t":"03-ingestion/pdfs-and-books","k":"mentions","w":4},{"s":"term:ingest","t":"09-maintenance/privacy-and-secrets","k":"mentions","w":4},{"s":"col:glossary","t":"term:wikilink","k":"contains","h":1},{"s":"term:wikilink","t":"track-graph/graph-stores","k":"mentions","w":2},{"s":"term:wikilink","t":"01-concepts/why-markdown-and-plain-text","k":"mentions"},{"s":"term:wikilink","t":"track-graph/build-extract","k":"mentions","w":2},{"s":"term:wikilink","t":"05-graphs/graph-basics","k":"mentions","w":2},{"s":"term:wikilink","t":"05-graphs/typed-links","k":"mentions","w":2},{"s":"term:wikilink","t":"09-maintenance/backups-and-portability","k":"mentions","w":2},{"s":"col:glossary","t":"term:backlinks","k":"contains","h":1},{"s":"term:backlinks","t":"04-structuring/naming-and-aliases","k":"mentions","w":2},{"s":"term:backlinks","t":"04-structuring/dedupe-and-merge","k":"mentions","w":2},{"s":"term:backlinks","t":"04-structuring/linking-rules","k":"mentions"},{"s":"term:backlinks","t":"05-graphs/metrics","k":"mentions"},{"s":"col:glossary","t":"term:atomic-pages","k":"contains","h":1},{"s":"term:atomic-pages","t":"04-structuring/atomic-pages","k":"mentions","w":2},{"s":"term:atomic-pages","t":"01-concepts/pkm-lineage","k":"mentions","w":2},{"s":"term:atomic-pages","t":"01-concepts/llm-wiki-pattern","k":"mentions"},{"s":"term:atomic-pages","t":"05-graphs/graph-vs-vectors","k":"mentions"},{"s":"col:glossary","t":"term:page-types","k":"contains","h":1},{"s":"term:page-types","t":"04-structuring/page-types","k":"mentions","w":6},{"s":"term:page-types","t":"04-structuring/atomic-pages","k":"mentions","w":5},{"s":"term:page-types","t":"03-ingestion/pdfs-and-books","k":"mentions","w":3},{"s":"term:page-types","t":"03-ingestion/voice-and-meetings","k":"mentions","w":3},{"s":"term:page-types","t":"03-ingestion/newsletters-and-email","k":"mentions","w":3},{"s":"term:page-types","t":"04-structuring/frontmatter-schema","k":"mentions","w":3},{"s":"term:page-types","t":"08-outputs/writing-from-the-vault","k":"mentions","w":3},{"s":"term:page-types","t":"10-troubleshooting/agent-writes-garbage","k":"mentions","w":3},{"s":"col:glossary","t":"term:frontmatter","k":"contains","h":1},{"s":"term:frontmatter","t":"04-structuring/frontmatter-schema","k":"mentions"},{"s":"term:frontmatter","t":"course-5-evals/evals-practice","k":"mentions","w":6},{"s":"term:frontmatter","t":"02-setup/obsidian-plugins","k":"mentions","w":3},{"s":"term:frontmatter","t":"03-ingestion/bulk-backfill","k":"mentions","w":2},{"s":"term:frontmatter","t":"05-graphs/typed-links","k":"mentions","w":2},{"s":"col:glossary","t":"term:aliases","k":"contains","h":1},{"s":"term:aliases","t":"04-structuring/naming-and-aliases","k":"mentions","w":10},{"s":"term:aliases","t":"04-structuring/dedupe-and-merge","k":"mentions","w":5},{"s":"term:aliases","t":"04-structuring/frontmatter-schema","k":"mentions","w":4},{"s":"term:aliases","t":"07-retrieval/rag-on-top","k":"mentions","w":2},{"s":"term:aliases","t":"10-troubleshooting/broken-links","k":"mentions","w":2},{"s":"col:glossary","t":"term:contradictions","k":"contains","h":1},{"s":"term:contradictions","t":"04-structuring/contradictions-and-supersession","k":"mentions","w":6},{"s":"term:contradictions","t":"07-retrieval/query-patterns","k":"mentions","w":3},{"s":"term:contradictions","t":"05-graphs/typed-links","k":"mentions","w":6},{"s":"term:contradictions","t":"track-graph/build-extract","k":"mentions","w":2},{"s":"term:contradictions","t":"04-structuring/dedupe-and-merge","k":"mentions","w":2},{"s":"term:contradictions","t":"06-agents/skills-and-commands","k":"mentions","w":2},{"s":"term:contradictions","t":"09-maintenance/review-cadence","k":"mentions","w":2},{"s":"col:glossary","t":"term:index-log","k":"contains","h":1},{"s":"term:index-log","t":"04-structuring/index-and-log","k":"mentions","w":2},{"s":"term:index-log","t":"03-ingestion/bulk-backfill","k":"mentions","w":3},{"s":"term:index-log","t":"07-retrieval/asking-questions","k":"mentions","w":3},{"s":"term:index-log","t":"02-setup/claude-code-setup","k":"mentions","w":2},{"s":"term:index-log","t":"02-setup/vault-structure","k":"mentions","w":2},{"s":"term:index-log","t":"04-structuring/contradictions-and-supersession","k":"mentions","w":2},{"s":"term:index-log","t":"06-agents/scheduled-maintenance","k":"mentions","w":2},{"s":"term:index-log","t":"06-agents/subagents","k":"mentions","w":2},{"s":"term:index-log","t":"09-maintenance/lint-and-health","k":"mentions","w":2},{"s":"term:index-log","t":"10-troubleshooting/broken-links","k":"mentions","w":2},{"s":"col:glossary","t":"term:knowledge-graph","k":"contains","h":1},{"s":"term:knowledge-graph","t":"track-graph/why-graphs","k":"mentions","w":2},{"s":"term:knowledge-graph","t":"05-graphs/obsidian-graph-view","k":"mentions","w":0},{"s":"term:knowledge-graph","t":"track-graph/resources","k":"mentions","w":5},{"s":"term:knowledge-graph","t":"track-graph/tools","k":"mentions","w":3},{"s":"term:knowledge-graph","t":"05-graphs/graph-basics","k":"mentions","w":2},{"s":"term:knowledge-graph","t":"05-graphs/graphrag","k":"mentions","w":2},{"s":"term:knowledge-graph","t":"05-graphs/exporting-your-graph","k":"mentions","w":2},{"s":"term:knowledge-graph","t":"10-troubleshooting/faq","k":"mentions","w":2},{"s":"res:5","t":"term:knowledge-graph","k":"related"},{"s":"col:glossary","t":"term:typed-links","k":"contains","h":1},{"s":"term:typed-links","t":"05-graphs/typed-links","k":"mentions"},{"s":"term:typed-links","t":"track-graph/why-graphs","k":"mentions","w":2},{"s":"term:typed-links","t":"track-graph/graph-stores","k":"mentions"},{"s":"term:typed-links","t":"05-graphs/obsidian-graph-view","k":"mentions"},{"s":"col:glossary","t":"term:orphans","k":"contains","h":1},{"s":"term:orphans","t":"05-graphs/metrics","k":"mentions","w":4},{"s":"term:orphans","t":"10-troubleshooting/broken-links","k":"mentions","w":3},{"s":"term:orphans","t":"track-graph/build-use","k":"mentions","w":2},{"s":"term:orphans","t":"05-graphs/obsidian-graph-view","k":"mentions","w":2},{"s":"term:orphans","t":"10-troubleshooting/vault-too-big","k":"mentions","w":2},{"s":"col:glossary","t":"term:hubs","k":"contains","h":1},{"s":"term:hubs","t":"04-structuring/linking-rules","k":"mentions","w":4},{"s":"term:hubs","t":"05-graphs/exporting-your-graph","k":"mentions","w":4},{"s":"term:hubs","t":"05-graphs/graph-basics","k":"mentions","w":5},{"s":"term:hubs","t":"05-graphs/obsidian-graph-view","k":"mentions","w":3},{"s":"term:hubs","t":"track-graph/build-use","k":"mentions","w":2},{"s":"term:hubs","t":"06-agents/agent-roles","k":"mentions","w":2},{"s":"term:hubs","t":"10-troubleshooting/common-failures","k":"mentions","w":2},{"s":"col:glossary","t":"term:clusters","k":"contains","h":1},{"s":"term:clusters","t":"track-evals/build-traces","k":"mentions","w":3},{"s":"term:clusters","t":"05-graphs/graphrag","k":"mentions","w":10},{"s":"term:clusters","t":"track-graph/build-query","k":"mentions","w":8},{"s":"term:clusters","t":"track-graph/graphrag","k":"mentions","w":7},{"s":"term:clusters","t":"track-graph/build-use","k":"mentions","w":6},{"s":"term:clusters","t":"05-graphs/obsidian-graph-view","k":"mentions","w":5},{"s":"term:clusters","t":"track-graph/tools","k":"mentions","w":3},{"s":"term:clusters","t":"02-setup/mcp-obsidian","k":"mentions","w":3},{"s":"term:clusters","t":"05-graphs/graph-basics","k":"mentions","w":3},{"s":"term:clusters","t":"track-graph/graph-stores","k":"mentions","w":2},{"s":"col:glossary","t":"term:graphrag","k":"contains","h":1},{"s":"term:graphrag","t":"track-graph/graphrag","k":"mentions","w":4},{"s":"term:graphrag","t":"05-graphs/graphrag","k":"mentions","w":2},{"s":"term:graphrag","t":"track-graph/resources","k":"mentions","w":6},{"s":"term:graphrag","t":"track-graph/tools","k":"mentions","w":4},{"s":"term:graphrag","t":"track-graph/build-query","k":"mentions","w":4},{"s":"term:graphrag","t":"track-graph/graph-stores","k":"mentions","w":2},{"s":"res:0","t":"term:graphrag","k":"related"},{"s":"res:66","t":"term:graphrag","k":"related"},{"s":"col:glossary","t":"term:rag","k":"contains","h":1},{"s":"term:rag","t":"01-concepts/wiki-vs-rag","k":"mentions","w":8},{"s":"term:rag","t":"track-graph/why-graphs","k":"mentions"},{"s":"term:rag","t":"07-retrieval/rag-on-top","k":"mentions","w":0},{"s":"term:rag","t":"track-graph/build-extract","k":"mentions","w":6},{"s":"term:rag","t":"track-graph/build-query","k":"mentions","w":5},{"s":"term:rag","t":"track-graph/resources","k":"mentions","w":5},{"s":"term:rag","t":"04-structuring/naming-and-aliases","k":"mentions","w":5},{"s":"term:rag","t":"track-graph/graphrag","k":"mentions","w":3},{"s":"res:0","t":"term:rag","k":"related"},{"s":"col:glossary","t":"term:vectors","k":"contains","h":1},{"s":"term:vectors","t":"05-graphs/graph-vs-vectors","k":"mentions","w":3},{"s":"term:vectors","t":"track-graph/why-graphs","k":"mentions","w":6},{"s":"term:vectors","t":"07-retrieval/rag-on-top","k":"mentions","w":6},{"s":"term:vectors","t":"01-concepts/wiki-vs-rag","k":"mentions","w":3},{"s":"col:glossary","t":"term:context-window","k":"contains","h":1},{"s":"term:context-window","t":"track-harness/build-guardrails","k":"mentions","w":3},{"s":"term:context-window","t":"track-harness/what-a-harness-is","k":"mentions","w":2},{"s":"term:context-window","t":"07-retrieval/context-budget","k":"mentions","w":0},{"s":"term:context-window","t":"07-retrieval/rag-on-top","k":"mentions","w":2},{"s":"col:glossary","t":"term:lost-in-middle","k":"contains","h":1},{"s":"term:lost-in-middle","t":"track-graph/build-query","k":"mentions","w":4},{"s":"term:lost-in-middle","t":"track-graph/build-extract","k":"mentions","w":2},{"s":"term:lost-in-middle","t":"course-1-context/how-models-read","k":"mentions"},{"s":"term:lost-in-middle","t":"track-graph/build-use","k":"mentions"},{"s":"res:2","t":"term:lost-in-middle","k":"related"},{"s":"col:glossary","t":"term:skills","k":"contains","h":1},{"s":"term:skills","t":"06-agents/skills-and-commands","k":"mentions","w":16},{"s":"term:skills","t":"track-harness/claude-code-as-harness","k":"mentions","w":4},{"s":"term:skills","t":"track-evals/tooling","k":"mentions","w":2},{"s":"term:skills","t":"10-troubleshooting/faq","k":"mentions","w":2},{"s":"res:96","t":"term:skills","k":"related"},{"s":"res:101","t":"term:skills","k":"related"},{"s":"res:108","t":"term:skills","k":"related"},{"s":"col:glossary","t":"term:slash-commands","k":"contains","h":1},{"s":"term:slash-commands","t":"track-graph/build-use","k":"mentions","w":2},{"s":"term:slash-commands","t":"06-agents/skills-and-commands","k":"mentions"},{"s":"term:slash-commands","t":"03-ingestion/web-clipper","k":"mentions"},{"s":"term:slash-commands","t":"03-ingestion/bulk-backfill","k":"mentions"},{"s":"col:glossary","t":"term:subagents","k":"contains","h":1},{"s":"term:subagents","t":"06-agents/subagents","k":"mentions","w":9},{"s":"term:subagents","t":"course-1-context/context-practice","k":"mentions","w":4},{"s":"term:subagents","t":"track-harness/claude-code-as-harness","k":"mentions","w":3},{"s":"term:subagents","t":"07-retrieval/context-budget","k":"mentions","w":2},{"s":"res:109","t":"term:subagents","k":"related"},{"s":"col:glossary","t":"term:hooks","k":"contains","h":1},{"s":"term:hooks","t":"06-agents/hooks","k":"mentions","w":12},{"s":"term:hooks","t":"course-4-harness/harness-practice","k":"mentions","w":10},{"s":"term:hooks","t":"track-harness/claude-code-as-harness","k":"mentions","w":8},{"s":"term:hooks","t":"course-4-harness/the-four-rings","k":"mentions","w":6},{"s":"term:hooks","t":"track-harness/build-graduate","k":"mentions","w":3},{"s":"term:hooks","t":"track-harness/context-engineering","k":"mentions","w":2},{"s":"col:glossary","t":"term:scheduled","k":"contains","h":1},{"s":"term:scheduled","t":"track-loop/build-overnight","k":"mentions","w":7},{"s":"term:scheduled","t":"track-loop/patterns","k":"mentions","w":2},{"s":"term:scheduled","t":"course-5-evals/evals-practice","k":"mentions"},{"s":"term:scheduled","t":"track-evals/build-ci","k":"mentions","w":9},{"s":"term:scheduled","t":"06-agents/scheduled-maintenance","k":"mentions","w":7},{"s":"term:scheduled","t":"course-2-loop/loop-practice","k":"mentions","w":4},{"s":"term:scheduled","t":"course-6-production/from-prototype","k":"mentions","w":4},{"s":"term:scheduled","t":"02-setup/claude-code-setup","k":"mentions","w":4},{"s":"term:scheduled","t":"06-agents/hooks","k":"mentions","w":3},{"s":"term:scheduled","t":"06-agents/skills-and-commands","k":"mentions","w":3},{"s":"col:glossary","t":"term:lint","k":"contains","h":1},{"s":"term:lint","t":"09-maintenance/lint-and-health","k":"mentions","w":3},{"s":"term:lint","t":"track-graph/build-use","k":"mentions","w":2},{"s":"term:lint","t":"01-concepts/llm-wiki-pattern","k":"mentions","w":3},{"s":"term:lint","t":"course-2-loop/the-four-parts","k":"mentions","w":2},{"s":"term:lint","t":"03-ingestion/bulk-backfill","k":"mentions","w":2},{"s":"term:lint","t":"10-troubleshooting/broken-links","k":"mentions","w":2},{"s":"res:57","t":"term:lint","k":"related"},{"s":"col:glossary","t":"term:git","k":"contains","h":1},{"s":"term:git","t":"02-setup/git-and-sync","k":"mentions","w":13},{"s":"term:git","t":"09-maintenance/versioning-with-git","k":"mentions","w":13},{"s":"term:git","t":"06-agents/safety-and-guardrails","k":"mentions","w":5},{"s":"term:git","t":"01-concepts/why-markdown-and-plain-text","k":"mentions"},{"s":"term:git","t":"track-evals/build-ci","k":"mentions","w":10},{"s":"term:git","t":"track-loop/build-overnight","k":"mentions","w":9},{"s":"term:git","t":"course-4-harness/harness-practice","k":"mentions","w":6},{"s":"term:git","t":"06-agents/scheduled-maintenance","k":"mentions","w":6},{"s":"term:git","t":"track-harness/build-guardrails","k":"mentions","w":5},{"s":"term:git","t":"02-setup/obsidian-plugins","k":"mentions","w":3},{"s":"res:54","t":"term:git","k":"related"},{"s":"col:glossary","t":"term:privacy","k":"contains","h":1},{"s":"term:privacy","t":"03-ingestion/chat-exports","k":"mentions","w":3},{"s":"term:privacy","t":"09-maintenance/privacy-and-secrets","k":"mentions","w":0},{"s":"term:privacy","t":"track-evals/build-ci","k":"mentions","w":2},{"s":"term:privacy","t":"02-setup/git-and-sync","k":"mentions","w":2},{"s":"term:privacy","t":"06-agents/safety-and-guardrails","k":"mentions","w":2},{"s":"col:glossary","t":"term:web-clipper","k":"contains","h":1},{"s":"term:web-clipper","t":"03-ingestion/web-clipper","k":"mentions","w":10},{"s":"term:web-clipper","t":"03-ingestion/newsletters-and-email","k":"mentions"},{"s":"term:web-clipper","t":"03-ingestion/bulk-backfill","k":"mentions"},{"s":"term:web-clipper","t":"06-agents/scheduled-maintenance","k":"mentions"},{"s":"res:12","t":"term:web-clipper","k":"related"},{"s":"col:glossary","t":"term:transcripts","k":"contains","h":1},{"s":"term:transcripts","t":"03-ingestion/youtube-transcripts","k":"mentions","w":10},{"s":"term:transcripts","t":"03-ingestion/voice-and-meetings","k":"mentions","w":4},{"s":"term:transcripts","t":"track-harness/context-engineering","k":"mentions","w":3},{"s":"term:transcripts","t":"track-harness/build-graduate","k":"mentions","w":3},{"s":"term:transcripts","t":"track-loop/context-hygiene","k":"mentions","w":3},{"s":"term:transcripts","t":"course-0-map/agents-or-workflows","k":"mentions","w":2},{"s":"term:transcripts","t":"course-1-context/the-four-places","k":"mentions","w":2},{"s":"term:transcripts","t":"course-4-harness/harness-practice","k":"mentions","w":2},{"s":"term:transcripts","t":"course-5-evals/two-kinds-of-checks","k":"mentions","w":2},{"s":"term:transcripts","t":"track-harness/build-guardrails","k":"mentions","w":2},{"s":"col:glossary","t":"term:agent","k":"contains","h":1},{"s":"term:agent","t":"course-0-map/what-an-agent-is","k":"mentions","w":13},{"s":"term:agent","t":"course-0-map/agents-or-workflows","k":"mentions","w":11},{"s":"term:agent","t":"course-6-production/operating-agents","k":"mentions","w":9},{"s":"term:agent","t":"course-6-production/day-one-plan","k":"mentions","w":9},{"s":"term:agent","t":"course-6-production/from-prototype","k":"mentions","w":7},{"s":"term:agent","t":"track-evals/agent-evals","k":"mentions","w":7},{"s":"term:agent","t":"01-concepts/why-markdown-and-plain-text","k":"mentions","w":5},{"s":"term:agent","t":"02-setup/git-and-sync","k":"mentions","w":5},{"s":"term:agent","t":"10-troubleshooting/common-failures","k":"mentions","w":3},{"s":"term:agent","t":"track-jev/jev-in-an-agent-stack","k":"mentions","w":2},{"s":"res:3","t":"term:agent","k":"related"},{"s":"res:4","t":"term:agent","k":"related"},{"s":"res:5","t":"term:agent","k":"related"},{"s":"res:8","t":"term:agent","k":"related"},{"s":"res:9","t":"term:agent","k":"related"},{"s":"res:10","t":"term:agent","k":"related"},{"s":"res:43","t":"term:agent","k":"related"},{"s":"res:82","t":"term:agent","k":"related"},{"s":"res:102","t":"term:agent","k":"related"},{"s":"res:108","t":"term:agent","k":"related"},{"s":"res:109","t":"term:agent","k":"related"},{"s":"res:110","t":"term:agent","k":"related"},{"s":"col:glossary","t":"term:five-layers","k":"contains","h":1},{"s":"term:five-layers","t":"course-0-map/five-layers","k":"mentions","w":2},{"s":"term:five-layers","t":"course-0-map/agents-or-workflows","k":"mentions","w":2},{"s":"term:five-layers","t":"course-6-production/day-one-plan","k":"mentions","w":2},{"s":"term:five-layers","t":"course-0-map/what-an-agent-is","k":"mentions"},{"s":"col:glossary","t":"term:workflows","k":"contains","h":1},{"s":"term:workflows","t":"course-0-map/agents-or-workflows","k":"mentions","w":7},{"s":"term:workflows","t":"course-2-loop/loops-vs-workflows","k":"mentions","w":7},{"s":"term:workflows","t":"course-0-map/what-an-agent-is","k":"mentions","w":3},{"s":"term:workflows","t":"track-evals/build-ci","k":"mentions","w":2},{"s":"term:workflows","t":"course-0-map/five-layers","k":"mentions","w":2},{"s":"term:workflows","t":"track-jev/what-jev-is-good-for","k":"mentions","w":2},{"s":"term:workflows","t":"06-agents/skills-and-commands","k":"mentions","w":2},{"s":"col:glossary","t":"term:context-engineering","k":"contains","h":1},{"s":"term:context-engineering","t":"course-1-context/context-practice","k":"mentions","w":3},{"s":"term:context-engineering","t":"track-harness/context-engineering","k":"mentions","w":2},{"s":"term:context-engineering","t":"course-1-context/the-four-places","k":"mentions"},{"s":"term:context-engineering","t":"track-harness/resources","k":"mentions","w":4},{"s":"term:context-engineering","t":"course-1-context/how-models-read","k":"mentions","w":2},{"s":"term:context-engineering","t":"track-loop/resources","k":"mentions","w":2},{"s":"col:glossary","t":"term:caching","k":"contains","h":1},{"s":"term:caching","t":"course-1-context/how-models-read","k":"mentions","w":11},{"s":"term:caching","t":"course-1-context/the-four-places","k":"mentions","w":6},{"s":"term:caching","t":"course-1-context/context-practice","k":"mentions","w":4},{"s":"term:caching","t":"course-6-production/operating-agents","k":"mentions","w":3},{"s":"term:caching","t":"course-6-production/from-prototype","k":"mentions","w":2},{"s":"term:caching","t":"track-harness/resources","k":"mentions","w":2},{"s":"col:glossary","t":"term:memory","k":"contains","h":1},{"s":"term:memory","t":"track-graph/tools","k":"mentions","w":6},{"s":"term:memory","t":"track-harness/context-engineering","k":"mentions","w":6},{"s":"term:memory","t":"track-loop/context-hygiene","k":"mentions","w":4},{"s":"term:memory","t":"track-harness/claude-code-as-harness","k":"mentions"},{"s":"term:memory","t":"track-graph/resources","k":"mentions","w":5},{"s":"term:memory","t":"course-1-context/the-four-places","k":"mentions","w":4},{"s":"term:memory","t":"track-graph/why-graphs","k":"mentions","w":3},{"s":"term:memory","t":"course-1-context/context-practice","k":"mentions","w":2},{"s":"term:memory","t":"course-4-harness/the-office","k":"mentions","w":2},{"s":"term:memory","t":"track-graph/building-graphs-with-llms","k":"mentions","w":2},{"s":"res:1","t":"term:memory","k":"related"},{"s":"res:7","t":"term:memory","k":"related"},{"s":"col:glossary","t":"term:loop","k":"contains","h":1},{"s":"term:loop","t":"track-loop/what-loop-engineering-is","k":"mentions","w":18},{"s":"term:loop","t":"course-2-loop/loop-practice","k":"mentions","w":15},{"s":"term:loop","t":"track-loop/patterns","k":"mentions","w":12},{"s":"term:loop","t":"track-loop/context-hygiene","k":"mentions","w":10},{"s":"term:loop","t":"course-2-loop/loops-vs-workflows","k":"mentions","w":9},{"s":"term:loop","t":"course-2-loop/the-four-parts","k":"mentions","w":9},{"s":"term:loop","t":"course-0-map/five-layers","k":"mentions","w":7},{"s":"term:loop","t":"course-0-map/agents-or-workflows","k":"mentions","w":6},{"s":"term:loop","t":"course-6-production/operating-agents","k":"mentions","w":4},{"s":"term:loop","t":"track-jev/jev-in-an-agent-stack","k":"mentions","w":4},{"s":"res:109","t":"term:loop","k":"related"},{"s":"col:glossary","t":"term:stop-conditions","k":"contains","h":1},{"s":"term:stop-conditions","t":"course-2-loop/the-four-parts","k":"mentions","w":11},{"s":"term:stop-conditions","t":"track-harness/build-guardrails","k":"mentions","w":6},{"s":"term:stop-conditions","t":"track-loop/stop-conditions","k":"mentions","w":4},{"s":"term:stop-conditions","t":"course-5-evals/two-kinds-of-checks","k":"mentions","w":2},{"s":"term:stop-conditions","t":"track-harness/context-engineering","k":"mentions","w":2},{"s":"term:stop-conditions","t":"07-retrieval/context-budget","k":"mentions"},{"s":"term:stop-conditions","t":"track-loop/what-loop-engineering-is","k":"mentions","w":5},{"s":"term:stop-conditions","t":"course-2-loop/loop-practice","k":"mentions","w":3},{"s":"term:stop-conditions","t":"track-loop/build-overnight","k":"mentions","w":3},{"s":"term:stop-conditions","t":"course-2-loop/loops-vs-workflows","k":"mentions","w":2},{"s":"col:glossary","t":"term:critics","k":"contains","h":1},{"s":"term:critics","t":"track-loop/critics-and-verification","k":"mentions","w":19},{"s":"term:critics","t":"track-loop/build-critic","k":"mentions","w":18},{"s":"term:critics","t":"course-2-loop/the-four-parts","k":"mentions","w":9},{"s":"term:critics","t":"track-loop/patterns","k":"mentions","w":12},{"s":"term:critics","t":"course-2-loop/loop-practice","k":"mentions","w":6},{"s":"term:critics","t":"track-loop/resources","k":"mentions","w":5},{"s":"term:critics","t":"track-loop/what-loop-engineering-is","k":"mentions","w":4},{"s":"term:critics","t":"track-loop/context-hygiene","k":"mentions","w":4},{"s":"term:critics","t":"track-loop/build-goal-test","k":"mentions","w":3},{"s":"term:critics","t":"track-loop/stop-conditions","k":"mentions","w":2},{"s":"col:glossary","t":"term:gate","k":"contains","h":1},{"s":"term:gate","t":"course-3-gate/gate-practice","k":"mentions","w":21},{"s":"term:gate","t":"course-3-gate/cheap-decisions","k":"mentions","w":15},{"s":"term:gate","t":"course-3-gate/gates-in-practice","k":"mentions","w":11},{"s":"term:gate","t":"course-0-map/five-layers","k":"mentions","w":5},{"s":"term:gate","t":"track-evals/build-ci","k":"mentions","w":5},{"s":"term:gate","t":"track-jev/build-router","k":"mentions","w":3},{"s":"term:gate","t":"track-harness/build-guardrails","k":"mentions","w":3},{"s":"term:gate","t":"course-6-production/day-one-plan","k":"mentions","w":2},{"s":"term:gate","t":"track-jev/jev-in-an-agent-stack","k":"mentions","w":3},{"s":"term:gate","t":"track-jev/build-swap-in-jev","k":"mentions","w":3},{"s":"col:glossary","t":"term:system-one","k":"contains","h":1},{"s":"term:system-one","t":"track-jev/resources","k":"mentions","w":15},{"s":"term:system-one","t":"track-jev/build-swap-in-jev","k":"mentions","w":12},{"s":"term:system-one","t":"track-jev/jev-in-an-agent-stack","k":"mentions","w":11},{"s":"term:system-one","t":"course-3-gate/gates-in-practice","k":"mentions","w":9},{"s":"term:system-one","t":"track-jev/what-jev-is-good-for","k":"mentions","w":8},{"s":"term:system-one","t":"track-jev/getting-started","k":"mentions","w":8},{"s":"term:system-one","t":"track-jev/system-one-models","k":"mentions","w":6},{"s":"term:system-one","t":"track-jev/build-decision-endpoint","k":"mentions","w":6},{"s":"term:system-one","t":"course-3-gate/cheap-decisions","k":"mentions","w":5},{"s":"term:system-one","t":"track-jev/build-router","k":"mentions","w":4},{"s":"col:glossary","t":"term:harness","k":"contains","h":1},{"s":"term:harness","t":"track-harness/what-a-harness-is","k":"mentions","w":14},{"s":"term:harness","t":"course-4-harness/the-office","k":"mentions","w":13},{"s":"term:harness","t":"track-harness/harness-landscape","k":"mentions","w":10},{"s":"term:harness","t":"course-4-harness/harness-practice","k":"mentions","w":9},{"s":"term:harness","t":"course-0-map/five-layers","k":"mentions","w":5},{"s":"term:harness","t":"track-loop/what-loop-engineering-is","k":"mentions","w":5},{"s":"term:harness","t":"course-6-production/day-one-plan","k":"mentions","w":4},{"s":"term:harness","t":"track-harness/claude-code-as-harness","k":"mentions","w":4},{"s":"term:harness","t":"track-evals/agent-evals","k":"mentions","w":2},{"s":"term:harness","t":"course-4-harness/the-four-rings","k":"mentions","w":5},{"s":"res:4","t":"term:harness","k":"related"},{"s":"col:glossary","t":"term:guardrails","k":"contains","h":1},{"s":"term:guardrails","t":"course-4-harness/the-four-rings","k":"mentions","w":7},{"s":"term:guardrails","t":"course-4-harness/harness-practice","k":"mentions","w":5},{"s":"term:guardrails","t":"06-agents/safety-and-guardrails","k":"mentions","w":4},{"s":"term:guardrails","t":"track-harness/what-a-harness-is","k":"mentions","w":3},{"s":"term:guardrails","t":"track-harness/claude-code-as-harness","k":"mentions","w":3},{"s":"term:guardrails","t":"track-harness/build-guardrails","k":"mentions","w":3},{"s":"term:guardrails","t":"02-setup/claude-code-setup","k":"mentions","w":2},{"s":"term:guardrails","t":"course-6-production/operating-agents","k":"mentions","w":3},{"s":"term:guardrails","t":"course-6-production/day-one-plan","k":"mentions","w":3},{"s":"term:guardrails","t":"track-harness/build-graduate","k":"mentions","w":3},{"s":"col:glossary","t":"term:tools-tool-use","k":"contains","h":1},{"s":"term:tools-tool-use","t":"track-harness/build-the-loop","k":"mentions","w":5},{"s":"term:tools-tool-use","t":"track-harness/build-guardrails","k":"mentions","w":4},{"s":"term:tools-tool-use","t":"course-6-production/operating-agents","k":"mentions","w":2},{"s":"term:tools-tool-use","t":"track-jev/jev-in-an-agent-stack","k":"mentions","w":2},{"s":"term:tools-tool-use","t":"track-harness/what-a-harness-is","k":"mentions","w":2},{"s":"term:tools-tool-use","t":"track-harness/build-graduate","k":"mentions","w":2},{"s":"col:glossary","t":"term:evals","k":"contains","h":1},{"s":"term:evals","t":"track-evals/tooling","k":"mentions","w":19},{"s":"term:evals","t":"track-evals/resources","k":"mentions","w":15},{"s":"term:evals","t":"course-0-map/five-layers","k":"mentions","w":7},{"s":"term:evals","t":"track-evals/why-evals","k":"mentions","w":7},{"s":"term:evals","t":"track-evals/agent-evals","k":"mentions","w":5},{"s":"term:evals","t":"track-evals/designing-evals","k":"mentions","w":3},{"s":"term:evals","t":"course-5-evals/evals-practice","k":"mentions","w":2},{"s":"term:evals","t":"course-6-production/day-one-plan","k":"mentions","w":2},{"s":"term:evals","t":"track-evals/build-ci","k":"mentions","w":9},{"s":"term:evals","t":"course-5-evals/two-kinds-of-checks","k":"mentions","w":4},{"s":"res:100","t":"term:evals","k":"related"},{"s":"res:111","t":"term:evals","k":"related"},{"s":"res:112","t":"term:evals","k":"related"},{"s":"col:glossary","t":"term:golden-set","k":"contains","h":1},{"s":"term:golden-set","t":"course-5-evals/judges-and-golden-sets","k":"mentions","w":6},{"s":"term:golden-set","t":"track-evals/designing-evals","k":"mentions","w":4},{"s":"term:golden-set","t":"track-evals/build-traces","k":"mentions","w":4},{"s":"term:golden-set","t":"track-evals/build-suite","k":"mentions","w":3},{"s":"term:golden-set","t":"track-evals/build-ci","k":"mentions","w":3},{"s":"col:glossary","t":"term:llm-judge","k":"contains","h":1},{"s":"term:llm-judge","t":"track-evals/llm-as-judge","k":"mentions","w":18},{"s":"term:llm-judge","t":"course-5-evals/judges-and-golden-sets","k":"mentions","w":13},{"s":"term:llm-judge","t":"track-evals/build-suite","k":"mentions","w":7},{"s":"term:llm-judge","t":"course-5-evals/evals-practice","k":"mentions","w":6},{"s":"term:llm-judge","t":"track-evals/resources","k":"mentions","w":5},{"s":"term:llm-judge","t":"course-3-gate/gate-practice","k":"mentions","w":4},{"s":"term:llm-judge","t":"track-loop/build-critic","k":"mentions","w":4},{"s":"term:llm-judge","t":"course-5-evals/two-kinds-of-checks","k":"mentions","w":3},{"s":"term:llm-judge","t":"track-loop/critics-and-verification","k":"mentions","w":3},{"s":"term:llm-judge","t":"course-2-loop/the-four-parts","k":"mentions","w":2},{"s":"res:10","t":"term:llm-judge","k":"related"},{"s":"col:glossary","t":"term:traces","k":"contains","h":1},{"s":"term:traces","t":"course-5-evals/evals-practice","k":"mentions","w":17},{"s":"term:traces","t":"track-evals/build-traces","k":"mentions","w":10},{"s":"term:traces","t":"course-5-evals/two-kinds-of-checks","k":"mentions","w":7},{"s":"term:traces","t":"course-6-production/operating-agents","k":"mentions","w":7},{"s":"term:traces","t":"track-evals/agent-evals","k":"mentions","w":4},{"s":"term:traces","t":"track-evals/tooling","k":"mentions","w":5},{"s":"term:traces","t":"track-evals/why-evals","k":"mentions","w":3},{"s":"term:traces","t":"track-evals/resources","k":"mentions","w":3},{"s":"term:traces","t":"course-6-production/from-prototype","k":"mentions","w":2},{"s":"col:glossary","t":"term:production","k":"contains","h":1},{"s":"term:production","t":"course-6-production/from-prototype","k":"mentions","w":9},{"s":"term:production","t":"course-5-evals/judges-and-golden-sets","k":"mentions","w":4},{"s":"term:production","t":"course-6-production/operating-agents","k":"mentions","w":5},{"s":"term:production","t":"track-harness/resources","k":"mentions","w":5},{"s":"term:production","t":"course-2-loop/loops-vs-workflows","k":"mentions","w":3},{"s":"term:production","t":"course-0-map/agents-or-workflows","k":"mentions","w":2},{"s":"term:production","t":"course-2-loop/the-four-parts","k":"mentions","w":2},{"s":"term:production","t":"course-2-loop/loop-practice","k":"mentions","w":2},{"s":"term:production","t":"track-jev/getting-started","k":"mentions","w":2},{"s":"term:production","t":"track-harness/harness-landscape","k":"mentions","w":2}]};