mcp-documentation-server — Claude Code skill for document search
The mcp-documentation-server skill indexes documents in a local database and searches by text or meaning. Use it to find relevant passages across files and retrieve their surrounding context.
no license 166 installs ★ 9 updated 2026-10-09
npx skills add reason-machines/mcp-skills --skill mcp-documentation-server -a claude-codeWhat mcp-documentation-server does
The skill manages a local document collection backed by the embedded Orama vector database. Search combines exact text matches with semantic results. Documents are split into parent and child chunks, so a result can be paired with its surrounding material.
Files placed in uploads can be processed in .txt, .md, and .pdf formats. Tools also let you add documents directly, list entries, retrieve full content, and delete records. A separate search tool covers the whole collection; another fetches nearby chunks for added context.
Gemini-generated search answers require GEMINI_API_KEY. Changing the embedding model means re-indexing every document, since vectors created by different models cannot be used together.
What it helps with
- find passages across a document collection by meaning
- locate exact text matches across all indexed documents
- process
.txt,.md, and.pdffiles fromuploads - retrieve nearby chunks to expand a search result
- ask Gemini to answer questions using selected documents
Example prompts for the agent
- Search all documents for the rate limiting configuration.
- Process the files in the uploads folder and add them to the collection.
- Use Gemini to answer my question about Redis from these documents.
How to install mcp-documentation-server
1Skills CLI (Vercel)
npx skills add reason-machines/mcp-skills --skill mcp-documentation-server -a claude-code2Manually: put the files in place
~/.claude/skills/mcp-documentation-server/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/mcp-documentation-server/
1Skills CLI (Vercel)
npx skills add reason-machines/mcp-skills --skill mcp-documentation-server -a codex2Manually: put the files in place
.agents/skills/mcp-documentation-server/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add reason-machines/mcp-skills --skill mcp-documentation-server -a cursor2Manually: put the files in place
.cursor/skills/mcp-documentation-server/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install reason-machines/mcp-skills skills/mcp-documentation-server2Manually: put the files in place
.github/skills/mcp-documentation-server/SKILL.md1Skills CLI (Vercel)
npx skills add reason-machines/mcp-skills --skill mcp-documentation-server -a gemini-cli2Platform CLI
gemini skills install https://github.com/reason-machines/mcp-skills --path skills/mcp-documentation-serverManually: put the files in place
.devin/skills/mcp-documentation-server/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add reason-machines/mcp-skills --skill mcp-documentation-server -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about mcp-documentation-server
How do I install mcp-documentation-server in Claude Code?
Run npx skills add reason-machines/mcp-skills --skill mcp-documentation-server -a claude-code. The skill file is at ~/.claude/skills/mcp-documentation-server/SKILL.md.
What does mcp-documentation-server do?
It stores and indexes documents locally, then searches them using text matches and semantic similarity. You can also retrieve full documents or nearby chunks for context.
Does Gemini search require an API key?
Yes. Gemini-generated search answers require GEMINI_API_KEY; hybrid text and vector search is available as a separate tool.