# Agent Harnesses MCP server

> Get opinionated AI agent harness picks, comparisons, decision guides and setup templates, with live GitHub and Hacker News discovery.

- Listing: https://mcp.tc/i/agent-harnesses
- Connect: this is a local (stdio) server; install it on your machine (see Install). The listing link is a page, not an MCP endpoint.
- Type: local (stdio)
- Auth: no API key needed
- Category: [AI & Machine Learning](https://mcp.tc/c/ai-ml)
- Vendor: RyanAlberts
- Homepage: <https://ryanalberts.github.io/best-of-Agent-Harnesses/>
- Docs: <https://github.com/RyanAlberts/best-of-Agent-Harnesses/tree/main/mcp>
- Repository: <https://github.com/RyanAlberts/best-of-Agent-Harnesses>
- Package: pypi `agent-harnesses-mcp`

## About

Serves the best-of-Agent-Harnesses list as MCP tools. An assistant can recommend a harness for a need, rank options for a use case, compare two to four projects side by side, search the list, and read decision guides, templates and playbooks. One tool also runs a live GitHub and Hacker News search for new, unvetted projects.

Runs locally over stdio as the PyPI package agent-harnesses-mcp, started with uvx (requires uv). It fetches harnesses.json from the GitHub repository at startup. No API key or sign-in is needed. Live discovery sends only the need text to api.github.com and hn.algolia.com.

## What it can do

- Get a single opinionated harness recommendation with alternatives and ones to avoid
- Rank harnesses for a use case by complexity, autonomy and recovery
- Compare 2 to 4 harnesses side by side
- Search the list by keyword, name, tag or category
- Read head-to-head decision guides and playbooks
- Fetch copy-paste setup templates such as AGENTS.md
- Discover fresh repos and Hacker News stories for a stack level

## Example prompts

- "Recommend an open-source harness for an always-on personal assistant in my chat apps"
- "Compare OpenClaw and Hermes agent side by side"
- "Which harnesses work best for sandboxed code execution of generated code?"
- "Give me the AGENTS.md template and write it into my project"

## Install

### Claude Code

1. Run this in a terminal, in your project folder:

```bash
claude mcp add --transport stdio agent-harnesses -- uvx agent-harnesses-mcp
```

2. Start Claude Code and type `/mcp`. **agent-harnesses** should show as connected.

Add `--scope user` to make it available in every project.

### Claude Desktop

1. Open **Settings → Developer → Edit Config**. It opens `claude_desktop_config.json`. Add:

`claude_desktop_config.json`:

```json
{
  "mcpServers": {
    "agent-harnesses": {
      "command": "uvx",
      "args": [
        "agent-harnesses-mcp"
      ]
    }
  }
}
```

2. Save the file and restart Claude Desktop.

Needs uv (Python) on your computer. The file lives in `~/Library/Application Support/Claude/` on macOS and `%APPDATA%\Claude\` on Windows.

### Cursor

[Add to Cursor](<https://cursor.com/install-mcp?name=agent-harnesses&config=eyJjb21tYW5kIjoidXZ4IiwiYXJncyI6WyJhZ2VudC1oYXJuZXNzZXMtbWNwIl19>) (opens Cursor)

Or add it by hand to `~/.cursor/mcp.json` (all projects) or `.cursor/mcp.json` (this project):

`mcp.json`:

```json
{
  "mcpServers": {
    "agent-harnesses": {
      "command": "uvx",
      "args": [
        "agent-harnesses-mcp"
      ]
    }
  }
}
```

Needs uv (Python) on your computer.

### VS Code

[Install in VS Code](<https://vscode.dev/redirect/mcp/install?name=agent-harnesses&config=%7B%22type%22%3A%22stdio%22%2C%22command%22%3A%22uvx%22%2C%22args%22%3A%5B%22agent-harnesses-mcp%22%5D%7D>) (opens VS Code)

Or from a terminal:

```bash
code --add-mcp '{"name":"agent-harnesses","type":"stdio","command":"uvx","args":["agent-harnesses-mcp"]}'
```

Or commit it to the repo in `.vscode/mcp.json`:

`.vscode/mcp.json`:

```json
{
  "servers": {
    "agent-harnesses": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "agent-harnesses-mcp"
      ]
    }
  }
}
```

Needs uv (Python) on your computer.

### Devin Desktop

1. Add it to `~/.config/devin/mcp_config.json` (macOS and Linux) or `%APPDATA%\devin\mcp_config.json` (Windows):

`mcp_config.json`:

```json
{
  "mcpServers": {
    "agent-harnesses": {
      "command": "uvx",
      "args": [
        "agent-harnesses-mcp"
      ]
    }
  }
}
```

2. Refresh the MCP server list in Cascade.

Devin Desktop is the new name for Windsurf.

### Codex

```bash
codex mcp add agent-harnesses -- uvx agent-harnesses-mcp
```

Or edit `~/.codex/config.toml` directly:

`config.toml`:

```toml
[mcp_servers.agent-harnesses]
command = "uvx"
args = ["agent-harnesses-mcp"]
```

Needs uv (Python) on your computer.

### Gemini CLI

```bash
gemini mcp add agent-harnesses uvx agent-harnesses-mcp
```

This adds it to the current project. Add `-s user` to use it everywhere.

### Any client

Most clients that start local servers accept this shape:

```json
{
  "mcpServers": {
    "agent-harnesses": {
      "command": "uvx",
      "args": [
        "agent-harnesses-mcp"
      ]
    }
  }
}
```

Zed puts servers under `context_servers` in its settings, with the same `command`, `args` and `env` fields.

Needs uv (Python) on your computer.

## Recent changes

- 2026-10-23 19:32 UTC: New version 0.8.0 (was 0.7.1), read from PyPI. Source: automatic check.
- 2026-10-09 07:27 UTC: New version 0.7.1 (was 0.6.0), read from PyPI. Source: automatic check.
- 2026-10-09 04:21 UTC: Tracking started: version 0.6.0. Source: automatic check.

More on the History tab, with descriptions before and after for the newest entries: https://mcp.tc/i/agent-harnesses#history (Atom feed: https://mcp.tc/i/agent-harnesses/history.xml). It records what mcp.tc's anonymous checks saw, every 3 hours where the tool list can be read without sign-in; it is not a security review.

## Details

- Server version: 0.8.0
- Last checked: 2026-10-09
- Listed: 2026-10-09
- Updated: 2026-10-09

---
Source: https://mcp.tc/i/agent-harnesses (mcp.tc is an independent directory, not affiliated with this server's publisher). Corrections: https://mcp.tc/report
