# compendio-mcp

Indexes project markdown files and serves relevant excerpts via local hybrid search. Keeps context clean by pulling single sections instead of full documents.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 5
- Author: RuloGB
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/RuloGB/compendio-mcp
- Updated: 2026-09-20
- HTML: https://affpapa.org/en/nevibecode/mcp/compendio-mcp

## What it does

Agents frequently burn context by reading entire source files with `grep` and `cat`. Compendio changes this habit. It combines keyword matching with semantic vector search, giving assistants direct access to the exact paragraph that answers the prompt.

The system stores data in a local SQLite file at `.compendio/compendio.db`. Vectors are calculated on your CPU with the `Xenova/multilingual-e5-small` model. No external network requests happen during retrieval, and running services or API tokens are unnecessary.

Running the tool requires Node.js version 22.12 or newer. The initial indexing step downloads an embedding model that takes tens of megabytes. Running `compendio index` in your terminal beforehand avoids timeouts on the first agent request.

## Requirements
- Needs: Node.js (commands run via npx)

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add compendio-mcp -- npx -y compendio-mcp
```

### Codex

**Platform CLI**
```bash
codex mcp add compendio-mcp -- npx -y compendio-mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "compendio-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "compendio-mcp"
            ]
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "compendio-mcp": {
            "type": "stdio",
            "command": "npx",
            "args": [
                "-y",
                "compendio-mcp"
            ]
        }
    }
}
```

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add compendio-mcp npx -y compendio-mcp
```

## When to use it
- reduce token consumption when querying local documentation
- retrieve project guidelines by conceptual meaning rather than exact phrasing
- prevent agents from reading entire files into memory
- sync documentation updates automatically without separate watcher processes

## Example prompts
- Search our markdown docs for the database migration workflow
- Locate the architecture decision record explaining token limits

## Similar
- [REA](https://affpapa.org/en/nevibecode/mcp/rea.md) — MCP server · ★ 392262
- [GitHub](https://affpapa.org/en/nevibecode/mcp/github.md) — MCP server · ★ 33989
- [Netdata](https://affpapa.org/en/nevibecode/mcp/netdata.md) — MCP server · ★ 81163
- [browser-use](https://affpapa.org/en/nevibecode/mcp/browser-use.md) — MCP server · ★ 119231
- [Codebase Memory](https://affpapa.org/en/nevibecode/mcp/codebase-memory.md) — MCP server · ★ 47577
- [mcp-server](https://affpapa.org/en/nevibecode/mcp/mcp-server.md) — MCP server · ★ 43902

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md