π The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
name: vector description: RuVector operations via npx [email protected] β embedding, search, RVF cognitive containers, GNN, attention, hooks, brain, sona, edge, identity
$ARGUMENTS
Vector operations via the ruvector npm package (pinned to 0.3.3). Parse subcommand from $ARGUMENTS.
Pinned version: [email protected] β every command below uses npx -y [email protected] ....
Usage: /vector [options]
Embedding
- embed
<text>βnpx -y [email protected] embed text "TEXT"384-dim ONNX vector. IfONNX WASM files not bundled, run/vector setup. - embed-adaptive
<text>βnpx -y [email protected] embed text "TEXT" --adaptive --domain codeLoRA-adapted embedding tuned to a domain. - embed-file
<path>β Read file, thennpx -y [email protected] embed text "$(cat <path>)" -o <path>.vec.json - embed-benchmark β
npx -y [email protected] embed benchmark(compares base vs adaptive).
Database lifecycle
- db create
<path>βnpx -y [email protected] create <path> -d 384 -m cosine - db stats
<path>βnpx -y [email protected] stats <path> - insert
<database><json-file>βnpx -y [email protected] insert <database> <json-file> - search
<database><vector-json>[-k N] βnpx -y [email protected] search <database> -v '[0.1,...]' -k N - export
<database>[-o file] [-f json|binary|parquet] [--compress] βnpx -y [email protected] export <database> -o backup.json - import
<file>[-d database] βnpx -y [email protected] import <file> -d <database> [--merge|--replace]
RVF (cognitive containers)
- rvf create
<path>βnpx -y [email protected] rvf create <path> - rvf ingest
<path>βnpx -y [email protected] rvf ingest <path> - rvf query
<path>βnpx -y [email protected] rvf query <path>(nearest neighbors) - rvf status
<path>βnpx -y [email protected] rvf status <path> - rvf segments
<path>βnpx -y [email protected] rvf segments <path> - rvf derive
<parent><child>βnpx -y [email protected] rvf derive <parent> <child>(lineage tracking) - rvf compact
<path>βnpx -y [email protected] rvf compact <path>(reclaim deleted space) - rvf examples β
npx -y [email protected] rvf examples(45 reference stores) - rvf download
<name>βnpx -y [email protected] rvf download <name>(e.g.agent_memory,swarm_knowledge)
GNN (Graph Neural Networks)
- gnn info β
npx -y [email protected] gnn info - gnn layer β
npx -y [email protected] gnn layer(build/test a multi-head attention GNN layer) - gnn search β
npx -y [email protected] gnn search(differentiable soft-attention search) - gnn compress β
npx -y [email protected] gnn compress(5-level adaptive tensor compression)
Attention mechanisms
- attention list β
npx -y [email protected] attention listLists ALL available mechanisms: DotProduct, MultiHead, Flash, Hyperbolic, Linear, MoE, GraphRoPe, EdgeFeatured, DualSpace, LocalGlobal. - attention compute β
npx -y [email protected] attention compute - attention benchmark β
npx -y [email protected] attention benchmark - attention hyperbolic β
npx -y [email protected] attention hyperbolic(Poincare-ball geometry ops; the real hyperbolic surface in 0.3.3) - attention info β
npx -y [email protected] attention info
Code intelligence (hooks)
- hooks init β
npx -y [email protected] hooks init --pretrain --build-agents quality - hooks stats β
npx -y [email protected] hooks stats(Q-learning patterns, vector memories, trajectories) - hooks route
<task>βnpx -y [email protected] hooks route "DESCRIPTION"(positional) - hooks route-enhanced
<task>βnpx -y [email protected] hooks route-enhanced "DESCRIPTION" - hooks suggest-context β
npx -y [email protected] hooks suggest-context - hooks ast-analyze
<file>βnpx -y [email protected] hooks ast-analyze <file>(positional) - hooks ast-complexity
<files...>βnpx -y [email protected] hooks ast-complexity <files...> - hooks diff-analyze [commit] β
npx -y [email protected] hooks diff-analyze HEAD - hooks diff-classify [commit] β
npx -y [email protected] hooks diff-classify HEAD - hooks diff-similar β
npx -y [email protected] hooks diff-similar - hooks coverage-route
<file>βnpx -y [email protected] hooks coverage-route <file> - hooks coverage-suggest
<files...>βnpx -y [email protected] hooks coverage-suggest <files...> - hooks graph-cluster
<files...>βnpx -y [email protected] hooks graph-cluster <files...>(spectral/Louvain) - hooks rag-context
<query>βnpx -y [email protected] hooks rag-context "QUERY" - hooks security-scan
<files...>βnpx -y [email protected] hooks security-scan <files...>(runhooks initfirst) - hooks remember
<content>βnpx -y [email protected] hooks remember "CONTENT" - hooks recall
<query>βnpx -y [email protected] hooks recall "QUERY" - hooks coedit-record / coedit-suggest β record + retrieve files edited together.
- hooks error-record / error-suggest β learn errorβfix pairs and retrieve.
- hooks trajectory-begin / trajectory-step / trajectory-end β record an execution trajectory.
- hooks pre-edit / post-edit / pre-command / post-command / session-start / session-end β Claude Code hook lifecycle.
Known bugs in 0.3.3 hooks:
force-learnreportsskipped: no trajectories to processuntil one is recorded;graph-mincutraisesCannot read properties of undefined;git-churnfails outside a git repo. Avoid these or run inside a git repo with seeded intelligence state.
Native + workers (background analysis)
- native list β
npx -y [email protected] native list(worker types: security, analysis, learning) - native run
<type>βnpx -y [email protected] native run security(or analysis|learning) - native benchmark β
npx -y [email protected] native benchmark - native compare β
npx -y [email protected] native compare - workers triggers / presets / phases / dispatch / status / stats / cleanup / cancel / run / create / init-config β all available via
npx -y [email protected] workers <subcmd>. First invocation auto-installsagentic-flow(slow).
Collective knowledge (brain) β needs @ruvector/pi-brain
- brain status β
npx -y [email protected] brain status - brain search
<query>βnpx -y [email protected] brain search "QUERY" - brain share
<title>βnpx -y [email protected] brain share "TITLE" - brain list / get / vote / delete / drift / partition / transfer / sync / page β full Brainpedia + LoRA-weight management.
SONA (Self-Optimizing Neural Architecture) β needs @ruvector/ruvllm
- sona status / info / stats β
npx -y [email protected] sona status - sona patterns
<query>βnpx -y [email protected] sona patterns "QUERY" - sona train
<data>βnpx -y [email protected] sona train <data>(record a training trajectory) - sona export β
npx -y [email protected] sona export(export learned weights)
LLM orchestration β needs @ruvector/ruvllm
- llm models β
npx -y [email protected] llm models - llm embed
<text>βnpx -y [email protected] llm embed "TEXT"(RuvLLM-backed embeddings) - llm benchmark β
npx -y [email protected] llm benchmark - llm info β
npx -y [email protected] llm info
Identity (pi key for brain/edge/MCP)
- identity generate β
npx -y [email protected] identity generate(creates a 64-hex-char pi key) - identity show β
npx -y [email protected] identity show - identity export
-o fileβnpx -y [email protected] identity export -o key.enc(encrypted backup) - identity import
<file>βnpx -y [email protected] identity import <file>
Edge compute network
- edge status β
npx -y [email protected] edge status - edge balance [nodeId] β
npx -y [email protected] edge balance(rUv balance) - edge tasks β
npx -y [email protected] edge tasks - edge join β
npx -y [email protected] edge join(join the network as a node) - edge dashboard β
npx -y [email protected] edge dashboard(opens in browser)
Server / Decompile / Demo
- server [-p 8080] [-g 50051] [-d data-dir] β
npx -y [email protected] server -p 8080 - decompile
<target>[-o dir] [-f modules|single|json] βnpx -y [email protected] decompile <npm-pkg-or-file-or-url> - demo --basic | --gnn | --graph β
npx -y [email protected] demo --basic(interactive tutorial)
System
- doctor β
npx -y [email protected] doctor(Node, npm, bindings, Rust) - info β
npx -y [email protected] info - benchmark β
npx -y [email protected] benchmark(known issue: fails withMissing field 'dimensions'on some installs; usegnn searchorattention benchmarkas alternatives) - install [pkg|--all] β
npx -y [email protected] install --all(lists/installs optional add-ons) - setup β
npx -y [email protected] setup(Setup Guide)
Setup helper
For first-run users hitting ONNX WASM files not bundled, Brain commands require @ruvector/pi-brain, or SONA not available, ask the user to run /ruflo-ruvector:vector-setup (it installs packages, so it is user-invoked: ADR-507).
MCP server (91 tools)
- Register once:
claude mcp add ruvector -- npx -y [email protected] mcp start - Verify:
claude mcp list | grep ruvector - Then call MCP tools directly (e.g.
hooks_route,hooks_ast_analyze,hooks_rag_context,brain_search,attention_list).
Not in 0.3.3 (do not invoke)
compare, top-level index, embed --file, embed --batch --glob, cluster --namespace --k (top-level cluster is "Coming Soon"), embed --model poincare, brain agi * (use brain status directly).