# Retrieve Local Documentation Efficiently

Finding relevant local documentation can waste time and context. This skill guides targeted retrieval, source consolidation, and web fallback when local coverage is insufficient.

## Install

```bash
npx skillstore add consiliency/docs-retrieval
```

## Metadata

- Status: approved
- Slug: consiliency-docs-retrieval
- Skillstore revision: r1
- Version status: missing
- Tree hash: f2b992909071d8c33b7dd62ee4d0147da653c323d9e386c5170a8097a9622241
- Author: Consiliency
- GitHub username: Consiliency
- License: MIT
- Repository: https://github.com/Consiliency/treesitter-chunker/tree/main/.ai-dev-kit/skills/docs-retrieval
- Ref: 635f69fb8d2f4e6330ba47a4e5a0fb239c04d110
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 79
- Quality tier: bronze
- Public page: https://skillstore.pages.dev/skills/consiliency-docs-retrieval
- Manifest: https://skillstore.pages.dev/api/skills/consiliency-docs-retrieval/manifest

## Capabilities

- Navigates hierarchical ai-docs indexes to locate relevant library pages.
- Selects targeted page summaries based on libraries, topics, and keywords.
- Consolidates findings from multiple libraries with sources and token estimates.
- Chooses focused summaries or full-context documentation based on task scope.
- Falls back to web search for missing or time-sensitive information.

## Use Cases

- Prepare feature implementation: Find focused library guidance, patterns, and gotchas before changing application code.
- Investigate dependency errors: Locate troubleshooting pages and known issues for the library that produced an error.
- Review cross-library architecture: Combine documentation from several libraries into one sourced context package.

## Prompt Templates

### Find one topic

```
Find local documentation for {library} about {topic}. Load the smallest relevant page set and summarize key points, gotchas, and source paths.
```

### Investigate an error

```
Use local ai-docs to investigate this {library} error: {error}. Retrieve troubleshooting guidance, likely causes, and source paths before proposing fixes.
```

### Compare multiple libraries

```
Gather targeted local documentation for {libraries} related to {feature}. Consolidate integration guidance, conflicts, gotchas, minimal patterns, sources, and estimated token use.
```

### Prepare implementation context

```
Build a documentation package for {task} across {libraries}. Prioritize local indexes, verify version-sensitive details online, identify uncertainties, and provide sourced implementation guidance.
```

## Limitations

- Requires a compatible local ai-docs hierarchy and index files.
- Does not create or update missing documentation.
- Local content may be stale, incomplete, or inconsistent with installed library versions.
- Full-context retrieval can consume 5,000 to 15,000 tokens.

## Best Practices

- Identify required libraries and topics before loading documentation.
- Use indexes and targeted page summaries before requesting full context.
- Record source paths and verify online when details are time-sensitive.

## Anti Patterns

- Do not load every library or full-context file without a defined need.
- Do not skip local documentation based on remembered API behavior.
- Do not make sub-agents repeat navigation when relevant context is already available.

## Security Audit

- Audited at: 2026-07-13T16:07:08.638\+00:00
- Summary: All 39 static findings are false positives caused by Markdown inline-code markers and fenced-code delimiters in SKILL.md. No shell execution, command interpolation, credential handling, malicious intent, or prompt injection was found.

## Stats

- Views: 191
- Downloads: 9
- Favorites: 0
- Popularity score: 0