hiivmind-corpus

Add documentation source to corpus. Triggers: "add source", "add git repo", "include blog posts", "add local documents", "extend corpus with web pages", "add team docs", "add PDF to corpus", "import PDF book", "split PDF into chapters".

Create and manage cross-corpus concept bridges in registry-graph.yaml. Use when users want to link concepts across corpora, create query-routing aliases, or validate cross-corpus relationships. Triggers: "bridge", "cross-corpus", "link corpora", "registry graph", "alias", "bridge show", "bridge validate", "add alias".

Non-interactive (re)build of a corpus index for pipelines. Replays the decisions captured in config.build (segmentation, organization, skip_sections, embeddings, verify prefs) instead of prompting, regenerates index.yaml + the rendered markdown, optional graph and embeddings, and writes build-result.yaml (headless contract). Triggers: "headless build", "rebuild corpus headless", "build result file", scheduled rebuilds. Users normally want `build`.

This skill should be used when the user asks to "build corpus index", "create index from docs", "analyze documentation", "populate corpus index", or needs to build the initial index for a corpus that was just initialized. Triggers on "build my corpus", "index the documentation", "create the index.md", "finish setting up corpus", "hiivmind-corpus build", or when a corpus has placeholder index.md that says "Run hiivmind-corpus-build", or "create the index.yaml".

This skill should be used when the user asks "what corpora do I have installed?", "list my documentation corpora", "find installed corpus plugins", "what hiivmind-corpus plugins are available?", "check the docs", "search the documentation", "what do the docs say", "look up in corpus", or needs to discover installed documentation corpus skills. Also use when a generic documentation query doesn't specify which corpus to search.

This skill should be used when the user asks to "add more detail to corpus", "enhance corpus coverage", "expand section in index", "need more depth on [topic]", "corpus is too shallow", "add entries for [topic]", or wants deeper coverage of specific topics in an existing corpus. Triggers on "enhance [topic] section", "more detail on [feature]", "expand the index", "add depth to corpus", or "hiivmind-corpus enhance".

Headless (non-interactive) enrichment of stale index entries. Re-scans changed source files to regenerate entry metadata (title, summary, tags, keywords, category), assigns concepts from the existing graph.yaml, verifies summaries against content, clears stale flags, re-renders, and re-embeds. Run after hiivmind-corpus-refresh-headless in automated pipelines, or standalone to repair a corpus with accumulated stale entries. Requires a v2 index (index.yaml).

View, validate, and edit a corpus concept graph (graph.yaml). Triggers: "graph", "show graph", "validate graph", "concept graph", "view concepts", "graph.yaml", "show relationships", "add concept", "add relationship".

This skill should be used when the user asks to "create a corpus", "initialize documentation", "set up docs for a library", "index this project's docs", "create documentation corpus", or mentions wanting to create a new documentation corpus. Also triggers on "new corpus", "corpus for [library name]", or "hiivmind-corpus init".

Headless v1 → v2 index migration. Parses entry lines from index.md and index-*.md, cross-references .source/ for metadata, emits index.yaml plus a render: block, renders deterministically, and diff-checks ID parity against the original. Writes migrate-result.yaml (headless contract). Triggers: "migrate corpus", "migrate to v2", "convert index.md to index.yaml".

Navigate and query documentation from registered corpora. Use when users ask documentation questions, want to look up API references, or mention keywords matching registered corpora. Triggers: documentation, docs, lookup, search corpus, what does, how to, API reference. Auto-triggers when query contains keywords from any registered corpus (flyio, polars, etc.).

Headless (non-interactive) variant of hiivmind-corpus-refresh for use by automated pipelines and scheduled tasks. Always runs in update mode, selects all stale sources automatically, and emits a machine-readable YAML result block instead of prose. Use hiivmind-corpus-refresh for interactive use.

This skill should be used when the user asks to "refresh corpus", "sync documentation", "update corpus index", "check for upstream changes", "corpus is stale", "docs are outdated", or mentions that documentation sources have changed. Triggers on "refresh my [corpus name] corpus", "sync corpus with upstream", "check if docs are current", "update from source repo", or "hiivmind-corpus refresh".

Register a documentation corpus with the current project. Use when users want to add a corpus to their project, connect documentation sources, or set up corpus access. Triggers: register corpus, add corpus, connect docs, add documentation, setup corpus.

Non-interactive corpus freshness check for pipelines. Compares upstream SHAs against config.yaml, counts stale entries and embedding lag, and writes status-result.yaml (headless contract). Cheap: no clones, no index edits. Triggers: "headless status", "status result file", scheduler pre-checks.

Check the health and status of registered corpora. Use when users want to see corpus freshness, verify configurations, check for upstream updates, or diagnose issues. Triggers: corpus status, check corpus, corpus health, is corpus up to date, corpus info.