# Tandem Docs

Tandem Docs connects an agent to the official Tandem documentation, so it answers from published pages instead of guessing commands and flags.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · 1205 installs
- Author: tandem
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://smithery.ai/server/tandem/docs-mcp
- Updated: 2026-04-02
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/mcp/tandem-docs

## What it does

Tandem Docs exposes the published Tandem docs through search, page retrieval, and structured learning guides. It helps when an agent installs Tandem Engine, writes code with the Python or TypeScript SDK, or builds workflows and automations: the model reads real product guidance rather than improvising.

The server ships thirteen tools. Search across the docs index, fetch a normalized page by path or URL, get a suggested starting point for a desktop user, server admin, or developer. Cache tools let you force-refresh the index or a single page and compare before/after hashes, so content changes are provable.

It runs remotely over HTTP. Nothing to install locally. The endpoint is `https://docs-mcp--tandem.run.tools`. When the network is down, cache tools report that data came from a stale fallback rather than the live source.

## Requirements
- Remote server: `https://docs-mcp--tandem.run.tools`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http docs-mcp https://docs-mcp--tandem.run.tools
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add tandem-docs --url https://docs-mcp--tandem.run.tools
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "tandem-docs": {
            "url": "https://docs-mcp--tandem.run.tools"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "tandem-docs": {
            "type": "http",
            "url": "https://docs-mcp--tandem.run.tools"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http tandem-docs https://docs-mcp--tandem.run.tools
```

## When to use it
- ground an agent in current Tandem documentation
- build a workflow from the official guide, not from model memory
- pick a docs starting point for a specific user role
- prove a docs page changed by comparing hashes
- warm the docs cache ahead of time for faster answers

## Example prompts
- Find the Tandem Engine server install steps in the docs.
- Point me to the right starting page for a developer using the Python SDK.
- Check whether the headless deployment page changed and show the diff.

## Similar
- [tandem-codex-plugin](https://affpapa.org/en/nevibecode/plugins/tandem-codex-plugin.md) — plugin · ★ 1596 repo stars
- [docs-mcp](https://affpapa.org/en/nevibecode/mcp/docs-mcp.md) — MCP server · ★ 125 repo stars
- [game-docs](https://affpapa.org/en/nevibecode/skills/game-docs.md) — skill · 2000 installs
- [gsd-ingest-docs](https://affpapa.org/en/nevibecode/skills/gsd-ingest-docs.md) — skill · 574 installs
- [project-docs](https://affpapa.org/en/nevibecode/skills/coco-research-project-docs.md) — skill · ★ 328 repo stars
- [absolute-docs](https://affpapa.org/en/nevibecode/skills/absolute-docs.md) — skill · 171 installs

---
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
