Skip to main content
Glama

Inkdex

npm

MCP server for semantic search over markdown docs.

Tools

Tool

Description

search_docs

Search indexed documentation. Returns matching chunks ranked by relevance.

Related MCP server: Markdown RAG Documentation

Usage

Add to your MCP client configuration:

{
  "mcpServers": {
    "inkdex": {
      "command": "npx",
      "args": [
        "-y",
        "inkdex"
      ],
      "env": {
        "DOCS_PATH": "/path/to/your/docs"
      }
    }
  }
}

Variable

Required

Description

DOCS_PATH

Yes

Path to the directory containing markdown files to index

How it works

Markdown files are split into chunks by heading structure and paragraph boundaries. Chunks are embedded locally using all-MiniLM-L6-v2 and stored in SQLite.

Search ranks results using both vector similarity and BM25 full-text matching, combined via Reciprocal Rank Fusion.

Results include the source, context, and text:

[mira/authentication.md]
<context>
Authentication > Token Expiration and Refresh
</context>
The refresh token expires...
A
license - permissive license
-
quality - not tested
A
maintenance

Maintenance

Maintainers
Response time
4dRelease cycle
4Releases (12mo)
Commit activity

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/antonlunden/inkdex'

If you have feedback or need assistance with the MCP directory API, please join our Discord server