an open-source alternative to the dots, bots and muses of the world, run an office of claude code/codex like agents on your laptop, sandboxes or anywhere, uses your existing subscriptions
Munder Difflin — Blog Idea Backlog (52 posts)
Author: Kevin (SEO strategist) · Pairs with
SEO_METADATA.md. Every idea maps to the keyword taxonomy and ladders up to a home pillar (#what= "what is a harness",#how= orchestration, MemPalace = memory). Slugs match the keyword→URL map inSEO_METADATA.md§1.2 — keep the slug so internal links and the sitemap line up.
Format note for Angela
The two tables below are the backlog. Schema for each row:
| Column | Meaning |
|---|---|
# |
Stable id (don't renumber; reference posts by id) |
Slug |
URL: https://munderdiffl.in/blog/<slug>/ — use verbatim |
Title |
Catchy, ≤60 chars where possible; becomes <h1> + <title> (append " — Munder Difflin") |
Type |
Technical or Non-technical (drives articleSection) |
Primary keyword |
Single focus keyword (one per post) |
Secondary keywords |
2–3 supporting terms |
Intent |
informational / how-to / comparison / commercial |
Audience |
Who it's for |
Synopsis |
1–2 line angle so a writer can execute without guessing |
Words |
Suggested length |
Internal links |
Pillar (↑) + sibling posts (→) to wire, by id or anchor |
Priority for the first batch: start with the P1 long-tail (#1, #3, #5, #9, #11) and P2
alternative/comparison (#27, #28, #29, #34, #35) rows — highest winnability for a new site. Each
post: TL;DR at top, one <h1>, descriptive H2/H3, link up to a pillar + 2–4 siblings + a
/#install download CTA. Add the URL to docs/sitemap.xml on publish.
Balance: 26 Technical · 26 Non-technical = 52 total.
A. Technical (26) — internals, how-tos, deep dives for builders
| # | Slug | Title | Type | Primary keyword | Secondary keywords | Intent | Audience | Synopsis | Words | Internal links |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | how-to-run-multiple-claude-code-agents | How to Run Multiple Claude Code Agents (Without Losing Track) | Technical | how to run multiple claude code agents | run multiple claude code agents, claude code parallel agents, claude code multi-agent setup | how-to | Claude Code users | Step-by-step: spin up several Claude Code sessions, give each a role, and coordinate them instead of alt-tabbing between windows. The flagship guide. | 1800 | ↑/#how · →#9, #11, #2, #27 |
| 2 | claude-code-orchestration-guide | Orchestrating Claude Code Agents: A Practical Guide | Technical | how to orchestrate claude code agents | claude code orchestration, claude code orchestrator, ai agent orchestration | how-to | Claude Code power users | What "orchestration" means for coding agents and how a GOD orchestrator routes, assigns, and escalates work across a hive. | 2000 | ↑/#how · →#1, #15, #20 |
| 3 | give-claude-code-long-term-memory | How to Give Claude Code Long-Term Memory | Technical | how to give claude code long-term memory | ai agent long-term memory, claude code memory, semantic memory for ai agents | how-to | Claude Code users | Why Claude forgets between sessions and how a markdown-first + semantic memory layer (MemPalace) makes agents recall across runs. | 1800 | ↑MemPalace(/#how) · →#16, #17, #5 |
| 4 | node-pty-electron-real-terminals | Real Terminals in Electron with node-pty | Technical | node-pty electron | node-pty, pseudo terminal, electron terminal app | how-to | Electron/JS devs | How we run byte-for-byte authentic shells inside Electron with node-pty, the native-rebuild gotcha, and streaming over IPC. | 2200 | ↑/#what · →#5, #6 |
| 5 | building-a-terminal-ui-xterm-node-pty | Building a Terminal UI with xterm.js and node-pty | Technical | xterm.js terminal app | xterm.js, terminal emulator, node-pty electron | how-to | Frontend/Electron devs | Wiring xterm.js to a node-pty backend: resize, write-back, performance, and rendering many live terminals at once. | 2200 | ↑/#what · →#4, #25 |
| 6 | single-committer-git-pattern | The Single-Committer Pattern: Multi-Agent Git Without Corruption | Technical | single committer git multi agent | git index.lock, multi-agent git, concurrent git writes | informational | Backend/tooling devs | Why parallel agents corrupt a git repo with index.lock races, and the single-committer mailbox design that fixes it. | 2000 | ↑/#what · →#7, #8 |
| 7 | atomic-file-mailboxes-for-agents | Atomic File Mailboxes: How Agents Message Each Other | Technical | can claude code agents talk to each other | ai agent messaging, agent mailbox, inter-agent communication | informational | Distributed-systems-curious devs | The outbox→router→inbox design that lets agents pass messages safely using plain files and atomic renames. | 1800 | ↑/#how · →#6, #20 |
| 8 | append-only-event-log-agents | An Append-Only Event Log for a Hive of Agents | Technical | ai agent event log | event sourcing, append-only log, agent observability | informational | Systems devs | How an append-only log makes a multi-agent system debuggable and replayable, and what we record per event. | 1800 | ↑/#how · →#6, #7 |
| 9 | manage-multiple-claude-code-sessions | How to Manage Multiple Claude Code Sessions Like a Pro | Technical | how to manage multiple claude code sessions | manage claude code sessions, claude code workflow, multiple terminals | how-to | Claude Code users | Tactics for juggling many sessions — naming, roles, context isolation — and when a harness beats manual tabs. | 1600 | ↑/#what · →#1, #11 |
| 10 | claude-code-hooks-explained | Claude Code Hooks, Explained (PreToolUse, PostToolUse, Stop) | Technical | claude code hooks | claude code hooks, pretooluse posttooluse, stop hook | informational | Claude Code devs | A practical tour of Claude Code's hook lifecycle and how we use a Unix-socket hook shim to drive the office floor. | 2000 | ↑/#how · →#15, #25 |
| 11 | claude-code-multi-agent-setup-tutorial | Set Up a Claude Code Multi-Agent Workflow in 10 Minutes | Technical | claude code multi-agent setup | claude code multi-agent setup, claude code parallel agents, getting started | how-to | New Munder Difflin users | A from-zero setup: install, onboarding wizard, spawn agents, and watch the GOD agent route the first task. | 1500 | ↑/#install · →#1, #9 |
| 12 | claude-code-git-worktrees-vs-hive | Git Worktrees vs a Hive: Two Ways to Parallelize Claude Code | Technical | claude code git worktrees | git worktrees, claude code parallel, isolated agent workspaces | comparison | Claude Code power users | When git worktrees are enough for parallel agents and when you need a coordinating hive on top. | 1800 | ↑/#how · →#1, #34 |
| 13 | human-in-the-loop-ai-agents | Human-in-the-Loop AI Agents: Approvals That Don't Slow You Down | Technical | human in the loop ai agents | human in the loop, ai approvals, agent guardrails | informational | AI/agent builders | Designing an approvals queue that only escalates the critical stuff (spend, destructive ops, scope) and stays autonomous otherwise. | 1900 | ↑/#how · →#20, #2 |
| 14 | visualizing-ai-agents-pixijs | Visualizing AI Agents with Pixi.js: An Office You Can Watch | Technical | ai agent visualization | ai agent visualization, pixijs, agent observability | informational | Frontend/creative devs | How we render agents as avatars on a Pixi.js office floor — pathfinding, seat assignment, flying message envelopes. | 2000 | ↑/#how · →#25, #5 |
| 15 | how-the-god-orchestrator-works | Inside the GOD Orchestrator: Routing, Adjudication, Escalation | Technical | claude code orchestrator | ai orchestrator, agent routing, supervisor agent | informational | Agent architects | A deep dive into the supervisor agent: how it reads requests, resolves routine ones, and escalates the rest. | 2200 | ↑/#how · →#2, #13, #7 |
| 16 | semantic-memory-for-ai-agents | Semantic Memory for AI Agents: Recall in Milliseconds | Technical | semantic memory for ai agents | semantic memory, vector recall, agent memory layer | informational | AI/agent builders | What semantic memory is, why markdown-first beats a heavyweight DB for agents, and how recall stays instant. | 2000 | ↑MemPalace(/#how) · →#3, #17 |
| 17 | markdown-first-agent-memory | Why We Built Agent Memory Markdown-First | Technical | ai agent long-term memory | markdown memory, agent knowledge base, long-term memory | informational | Tooling devs | The case for plain-markdown memory with a semantic index on top: human-readable, diffable, and degrades gracefully. | 1700 | ↑MemPalace(/#how) · →#3, #16 |
| 18 | local-first-ai-agent-orchestration | Local-First AI Agent Orchestration: Your Machine, Your Agents | Technical | local-first ai agent orchestration | local-first ai, local ai dev environment, privacy | informational | Privacy-minded devs | Why running the harness, agents, and memory locally beats cloud orchestration on control, cost, and privacy. | 1800 | ↑/#why · →#19, #36 |
| 19 | architecture-two-planes-one-renderer | Munder Difflin's Architecture: Two Data Planes, One Renderer | Technical | multi-agent harness architecture | electron architecture, ipc, agent harness design | informational | Senior devs | A walkthrough of the terminal plane (node-pty) and event plane (hooks/hive) feeding one React+Pixi renderer. | 2200 | ↑/#what · →#4, #14, #15 |
| 20 | coordinating-ai-coding-agents | The Best Way to Coordinate AI Coding Agents | Technical | best way to coordinate ai coding agents | coordinate ai agents, ai coding agent teams, agent collaboration | informational | Agent builders | Patterns for getting multiple coding agents to cooperate instead of clobbering each other's work. | 1900 | ↑/#how · →#2, #7, #15 |
| 21 | claude-code-subagents-vs-multi-agent-harness | Claude Code Subagents vs a Multi-Agent Harness | Technical | claude code subagents | claude code subagents, multi-agent harness, subagent limits | comparison | Claude Code users | Where built-in subagents stop and an external harness with memory + messaging takes over. | 1700 | ↑/#what · →#1, #34 |
| 22 | claude-code-automation-while-you-sleep | Claude Code Automation: Let Agents Build While You Sleep | Technical | claude code automation | claude code automation, overnight ai coding, autonomous agents | how-to | Claude Code users | How to set up long-running autonomous agents that keep working overnight, with guardrails so you wake up to progress not chaos. | 1800 | ↑/#why · →#40, #13 |
| 23 | debugging-multi-agent-systems | How to Debug a Multi-Agent System | Technical | debug multi-agent system | multi-agent debugging, agent observability, event log | how-to | Agent builders | Using the event log, per-agent terminals, and message trails to find why a hive went sideways. | 1800 | ↑/#how · →#8, #14 |
| 24 | mcp-and-skills-in-a-hive | Bringing Your MCP Servers and Skills Into a Hive of Agents | Technical | claude code mcp multi-agent | mcp servers, claude code skills, agent tools | how-to | Claude Code devs | How agents in the hive inherit your existing MCP servers, skills, and tools — and how to scope them per agent. | 1700 | ↑/#claude · →#11, #2 |
| 25 | rendering-many-live-terminals-performance | Rendering Many Live Terminals Without Melting the CPU | Technical | xterm.js performance | terminal rendering, xterm.js performance, electron performance | how-to | Frontend/Electron devs | Performance techniques for streaming and rendering dozens of live PTY terminals in one window. | 1900 | ↑/#what · →#5, #14 |
| 26 | building-an-ai-office-floor | Building an AI "Office Floor": Avatars, Pathfinding, Envelopes | Technical | ai agent visualization | tiled maps, pathfinding, game-ui for tools | informational | Creative/game-tooling devs | The game-dev techniques (Tiled maps, A* pathfinding, sprite recoloring) behind a developer tool's office floor. | 2000 | ↑/#how · →#14, #25 |
B. Non-technical (26) — concepts, comparisons, use-cases, opinion, AEO explainers
| # | Slug | Title | Type | Primary keyword | Secondary keywords | Intent | Audience | Synopsis | Words | Internal links |
|---|---|---|---|---|---|---|---|---|---|---|
| 27 | what-is-a-multi-agent-harness | What Is a Multi-Agent Harness? (Plain-English Guide) | Non-technical | claude code multi-agent | multi-agent harness, multi-agent ai framework, ai agent harness | informational | Devs new to agents | The canonical AEO explainer: define a multi-agent harness in one sentence up top, then how it differs from a single agent or a framework. | 1500 | ↑/#what · →#1, #2, #28 |
| 28 | what-are-claude-code-agents | What Are Claude Code Agents? And How to Use Many at Once | Non-technical | claude code agents | claude code agents, claude code subagents, ai coding agents | informational | Claude Code beginners | Explains agents in Claude Code and the leap from one agent to a coordinated team. | 1400 | ↑/#what · →#1, #21 |
| 29 | best-claude-code-multi-agent-tools | The Best Tools to Run Multiple Claude Code Agents (2026) | Non-technical | best claude code multi-agent tools | claude code multi-agent tool, agentic coding tools, best tools to run multiple claude code agents | commercial | Tool shoppers | An honest roundup of the multi-agent Claude Code landscape (Claude Squad, Conductor, Crystal, vibe-kanban, Munder Difflin) with a comparison table. | 2200 | ↑/#what · →#30, #31, #32, #33 |
| 30 | claude-squad-alternative | Looking for a Claude Squad Alternative? | Non-technical | claude squad alternative | claude squad alternative, claude code multi-agent tool, claude squad vs munder difflin | commercial | Claude Squad users | What Claude Squad does well, where it stops, and why a memory-backed visual hive might fit better. | 1500 | ↑/#what · →#29, #34 |
| 31 | conductor-claude-code-alternative | A Conductor Alternative for Orchestrating Claude Code | Non-technical | conductor claude code alternative | conductor claude code alternative, claude code orchestration tools | commercial | Conductor users | Honest comparison of Conductor's approach vs a local hive with a GOD orchestrator + shared memory. | 1500 | ↑/#how · →#29, #35 |
| 32 | crystal-claude-code-alternative | A Crystal Alternative: Multi-Agent Claude Code with Memory | Non-technical | crystal claude code alternative | crystal claude code alternative, claude code multi-agent tool | commercial | Crystal users | Where Crystal fits and what changes when agents share long-term memory and a visual floor. | 1400 | ↑/#what · →#29, #34 |
| 33 | vibe-kanban-alternative | A vibe-kanban Alternative for Coordinating AI Agents | Non-technical | vibe kanban alternative | vibe kanban alternative, ai agent task board, agentic coding tools | commercial | vibe-kanban users | Kanban-for-agents vs a self-routing hive — when you want a board and when you want an orchestrator. | 1400 | ↑/#how · →#29, #20 |
| 34 | claude-squad-vs-munder-difflin | Claude Squad vs Munder Difflin: Which Should You Use? | Non-technical | claude squad vs munder difflin | claude squad vs munder difflin, claude code multi-agent tool | comparison | Tool shoppers | Head-to-head: terminal-multiplexing vs a memory-backed, visual, orchestrated hive. Feature table + verdict. | 1600 | ↑/#what · →#29, #30 |
| 35 | claude-code-orchestration-tools-compared | Claude Code Orchestration Tools, Compared | Non-technical | claude code orchestration tools compared | claude code orchestration tools compared, best claude code multi-agent tools | comparison | Tool shoppers | A criteria-based comparison (memory, messaging, visibility, control, local-first) across the orchestration tools. | 1900 | ↑/#how · →#29, #31 |
| 36 | why-local-first-matters-for-ai-agents | Why Local-First Matters for Your AI Agents | Non-technical | local ai dev environment | local-first ai, privacy, ai data control | informational | Privacy-minded devs | The control/privacy/cost case for keeping agents and their memory on your own machine. | 1400 | ↑/#why · →#18, #37 |
| 37 | run-an-office-of-ai-agents | Run an Office of AI Agents While You Sleep | Non-technical | autonomous software team ai | ai pair programming team, ai coding agent teams, run agents while you sleep | informational | Indie devs / small teams | The vision post: a self-coordinating "office" of agents that keeps shipping after you log off. | 1500 | ↑/#why · →#22, #40 |
| 38 | can-claude-code-agents-talk-to-each-other | Can Claude Code Agents Talk to Each Other? | Non-technical | can claude code agents talk to each other | inter-agent messaging, ai agent communication, claude code agents | informational | Curious Claude Code users | Direct AEO answer ("Yes — here's how") then a plain-English tour of mailboxes and the router. | 1200 | ↑/#how · →#7, #28 |
| 39 | what-is-aeo-for-dev-tools | AEO for Dev Tools: Getting Cited by ChatGPT and Claude | Non-technical | answer engine optimization | aeo, generative engine optimization, ai search | informational | Indie hackers / DevRel | How we optimize Munder Difflin to be cited by AI answer engines, and what any dev tool can copy. | 1600 | ↑/#what · →#27, #44 |
| 40 | overnight-ai-coding-realistic-expectations | Overnight AI Coding: What's Realistic (and What's Not) | Non-technical | overnight ai coding | autonomous coding agents, overnight ai coding, ai coding limits | informational | Skeptical devs | An honest take on letting agents run unattended — what works today, where guardrails matter, what still needs you. | 1600 | ↑/#why · →#22, #13 |
| 41 | the-office-parody-behind-munder-difflin | The Office Parody Behind Munder Difflin (and Why It Helps) | Non-technical | munder difflin | munder difflin app, munder difflin claude code | informational | Fans / new visitors | The brand story — why "the world's best agents, the world's worst paper company" is more than a joke. | 1100 | ↑/ · →#37, #42 |
| 42 | why-we-built-munder-difflin | Why We Built Munder Difflin | Non-technical | munder difflin | munder difflin github, multi-agent harness | informational | Prospective users / contributors | Founder's-note origin story: the frustration with juggling terminals that led to a coordinated hive. | 1300 | ↑/ · →#41, #45 |
| 43 | how-to-choose-a-multi-agent-tool | How to Choose a Multi-Agent Coding Tool: A Buyer's Checklist | Non-technical | agentic coding tools | agentic coding tools, ai coding agent teams, choosing ai tools | commercial | Tool shoppers | A decision framework (memory, control, visibility, cost, local-first) for picking an agent tool, with a scoring rubric. | 1700 | ↑/#what · →#29, #35 |
| 44 | ai-agent-glossary | The AI Coding Agent Glossary: Harness, Orchestrator, Hive, Subagent | Non-technical | multi-agent ai framework | ai agent glossary, orchestrator vs harness, agentic terms | informational | Devs new to agents | Snackable AEO-friendly definitions of the terms everyone trips over. Each term = a quotable one-liner. | 1400 | ↑/#what · →#27, #39 |
| 45 | open-source-ai-tools-on-purpose | Building AI Dev Tools in the Open, On Purpose | Non-technical | open source ai agents | open source ai tools, mit license, building in public | informational | OSS-minded devs | Why Munder Difflin is MIT and built in public, and what that means for users and contributors. | 1300 | ↑/#opensource · →#42, #46 |
| 46 | contributing-to-munder-difflin | First-Time Contributor's Guide to Munder Difflin | Non-technical | munder difflin github | contributing, open source contribution, good first issue | informational | Potential contributors | A friendly path into the codebase: good first areas, how to run it, and how PRs work. | 1400 | ↑/#opensource · →#45, #19 |
| 47 | do-more-with-less-model-routing | Do More With Less: Smart Model Routing Across Agents | Non-technical | ai model routing | model routing, llm cost optimization, sonnet vs opus | informational | Cost-conscious devs | The cost argument: a lean orchestrator routing to heavier specialists only when needed — right model, right task. | 1500 | ↑/#why · →#15, #2 |
| 48 | from-one-terminal-to-a-team | From One Terminal to a Team: Scaling Your Claude Code Workflow | Non-technical | run multiple claude code agents | claude code workflow, scaling ai agents, ai coding agent teams | informational | Solo devs scaling up | A narrative of the workflow shift from a single session to a coordinated team, with concrete before/after. | 1500 | ↑/#what · →#1, #9 |
| 49 | ai-agents-that-actually-remember | AI Agents That Actually Remember: The Memory Problem | Non-technical | ai agent long-term memory | ai agent memory, semantic memory for ai agents, context loss | informational | AI-curious devs | Plain-English framing of why agents forget and what shared long-term memory changes day to day. | 1400 | ↑MemPalace(/#how) · →#3, #16 |
| 50 | watching-ai-agents-work-is-useful | Yes, Watching AI Agents Work Is Actually Useful | Non-technical | ai agent visualization | ai agent observability, watching agents, agent transparency | informational | Skeptical devs / managers | The case that a visual office floor isn't a gimmick — visibility builds trust and catches problems early. | 1300 | ↑/#how · →#14, #13 |
| 51 | small-team-ai-agents-playbook | A Small Team's Playbook for AI Coding Agents | Non-technical | ai coding agent teams | small team ai, ai pair programming team, team workflow | informational | Small dev teams / leads | How a 2–5 person team can adopt a hive of agents without losing control or review discipline. | 1600 | ↑/#why · →#37, #43 |
| 52 | munder-difflin-faq | Munder Difflin FAQ: Everything People Ask | Non-technical | munder difflin app | munder difflin download, munder difflin claude code, is munder difflin free | informational | New visitors | An FAQ-schema'd page answering the top questions (what is it, free?, local?, platforms, vs N terminals) — pure AEO play. | 1200 | ↑/ · →#27, #11, #41 |
52 ideas · 26 Technical · 26 Non-technical · every slug aligned to SEO_METADATA.md §1.2. Questions / re-prioritization → god (conversation: munder-blog-seo).