# ref-tools-ref-tools-mcp

Finds up-to-date technical documentation for APIs and libraries while trimming irrelevant sections to save tokens.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed · ★ 1183
- Author: ref-tools
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/ref-tools/ref-tools-mcp
- Updated: 2026-10-24
- HTML: https://affpapa.org/en/nevibecode/mcp/ref-tools-ref-tools-mcp

## What it does

The server exposes two tools to your coding agent: `ref_search_documentation` for querying tech docs and `ref_read_url` to fetch and parse pages as Markdown. Instead of pulling an entire web page into context, it analyzes earlier queries from the session to drop unneeded sections.

Ref avoids returning duplicate links during follow-up searches within the same session. Without a valid `REF_API_KEY`, the server cannot execute search queries.

## Requirements
- Remote server: `https://server.smithery.ai/@ref-tools/ref-tools-mcp/mcp`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http ref-tools-ref-tools-mcp https://server.smithery.ai/@ref-tools/ref-tools-mcp/mcp
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add ref-tools-ref-tools-mcp --url https://server.smithery.ai/@ref-tools/ref-tools-mcp/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "ref-tools-ref-tools-mcp": {
            "url": "https://server.smithery.ai/@ref-tools/ref-tools-mcp/mcp"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "ref-tools-ref-tools-mcp": {
            "type": "http",
            "url": "https://server.smithery.ai/@ref-tools/ref-tools-mcp/mcp"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http ref-tools-ref-tools-mcp https://server.smithery.ai/@ref-tools/ref-tools-mcp/mcp
```

## When to use it
- search public and private technical documentation directly from an agent
- strip noise from lengthy doc pages to avoid wasting context tokens
- prevent duplicate search results during multi-step prompt refinements
- convert web documentation pages to Markdown for model consumption

## Example prompts
- Search for Figma Comment API documentation using `ref_search_documentation`
- Fetch https://docs.n8n.io with `ref_read_url` and extract node details
- Query technical docs to compare merge and code nodes in n8n

## Similar
- [REA](https://affpapa.org/en/nevibecode/mcp/rea.md) — MCP server · ★ 386989
- [GitHub](https://affpapa.org/en/nevibecode/mcp/github.md) — MCP server · ★ 33918
- [Netdata](https://affpapa.org/en/nevibecode/mcp/netdata.md) — MCP server · ★ 81115
- [browser-use](https://affpapa.org/en/nevibecode/mcp/browser-use.md) — MCP server · ★ 119160
- [Codebase Memory](https://affpapa.org/en/nevibecode/mcp/codebase-memory.md) — MCP server · ★ 47392
- [mcp-server](https://affpapa.org/en/nevibecode/mcp/mcp-server.md) — MCP server · ★ 43872

---
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