Skip to main content
Glama

How to connect Lichess to Claude Code, ChatGPT and any AI agent

Yes, Claude, ChatGPT and other AI agents can work in your Lichess account. Cotlet.dev connects your own signed-in Lichess account to any AI agent, so you can ask in plain words and the AI accepts team join requests and sends updates to your team for you.

What you can ask

Accept the join requests from tom_k and anna_plays to my "Riverside Chess Club" team.

Show me all pending join requests for my team and who sent them this week.

Send a message to all members of "Riverside Chess Club": our blitz arena starts Thursday at 7 pm.

Accept every join request from today, then tell the team that 4 new players joined.

Message my team that Saturday's lesson with coach Maria moves to 10 am.

Related MCP server: LinkedIn MCP Server

Connect in 2 minutes

Get your API key at https://cotlet.dev/app and export it as COTLET_API_KEY.

Claude Code

claude mcp add --transport http cotlet https://cotlet.dev/mcp/v1 --header "Authorization: Bearer $COTLET_API_KEY"

ChatGPT

Settings → Apps → add connector URL https://cotlet.dev/mcp/v1, then sign in with Cotlet.dev. No API key needed.

Cursor

Copy .cursor/mcp.json from this repo into your project:

{
  "mcpServers": {
    "cotlet": {
      "url": "https://cotlet.dev/mcp/v1",
      "headers": {
        "Authorization": "Bearer ${COTLET_API_KEY}"
      }
    }
  }
}

Run it locally (stdio)

This repo is a real MCP server. It speaks MCP over stdio, lists its Lichess tools without any key, and forwards every tool call to the hosted Cotlet.dev server. The hosted URL above stays the recommended way. Use this when your client only starts local servers.

{"mcpServers":{"lichess":{"command":"uvx","args":["lichess-mcp"],"env":{"COTLET_API_KEY":"your-key"}}}}

Or with Docker:

docker build -t lichess-mcp .
docker run -i --rm -e COTLET_API_KEY lichess-mcp

Tools: lichess_guide, lichess_status, lichess_open, lichess_read, lichess_click, lichess_type, lichess_close. Without COTLET_API_KEY a tool call answers: "Set COTLET_API_KEY. Get a key at https://cotlet.dev/app".

Command line

Python 3.10+, nothing else to install. From PyPI:

pip install lichess-mcp        # or run it once: uvx lichess-mcp connect
lichess-mcp connect

Or straight from this repo:

export COTLET_API_KEY="your-key"
python3 lichess_agent.py connect   # ready config for Claude Code, Cursor, ChatGPT
python3 lichess_agent.py status    # is Lichess ready in your Cotlet.dev account
python3 lichess_agent.py guide     # Lichess screens and sign-in link
python3 lichess_agent.py read      # open Lichess, read the page, close the browser
python3 lichess_agent.py tools     # list Cotlet.dev tools
python3 lichess_agent.py serve     # start the stdio MCP server (same as no argument)

Questions from Lichess users

Does Lichess have an API?

Lichess has a full, free REST API with personal and OAuth tokens, but no official MCP server (checked 17 Oct 2026). The API covers teams, Arena and Swiss tournaments, studies, broadcasts, challenges and game export, so the gap is a hosted, maintained MCP with no token to create, scope or revoke. Cotlet.dev works in your own signed-in Lichess account, so your AI agent can accept team join requests and message your team with no API key and no code.

How do I connect Lichess to ChatGPT?

Add Cotlet.dev as an app in ChatGPT (Settings → Apps, connector URL https://cotlet.dev/mcp/v1), sign in to Lichess once in your Cotlet.dev Cloud Browser, and ask ChatGPT in plain words. No code and no Lichess API key.

Is there a way to use Claude with Lichess?

Yes. In Claude Code run the one-line claude mcp add command above; in the Claude app add https://cotlet.dev/mcp/v1 as a custom connector. Claude then works in your own Lichess account.

Is there a Lichess MCP server?

We found no official Lichess app in the ChatGPT or Claude directories. Cotlet.dev is a hosted MCP server that works in Lichess for you, 24/7, with Claude, ChatGPT, Cursor, Codex, n8n, Make and Zapier.

Find it on

Also works with


More: https://cotlet.dev/integrations/lichess-api-mcp-server?utm_source=github&utm_medium=readme&utm_campaign=lichess

Start free: https://cotlet.dev/signup?app=lichess&utm_source=github&utm_medium=readme&utm_campaign=lichess

Available Tools

7 tools
lichess_clickBInspect

Click a numbered element on the current Lichess page.

Inputs: element (required).

lichess_closeAInspect

Close the Lichess browser session when the work is done, so credits stop running.

Inputs: no named parameters.

lichess_guideAInspect

How to work in Lichess (accept team join requests and message your team): its screens and the sign-in link. Call this first.

Inputs: task.

lichess_openBInspect

Open a Lichess page in your signed-in Cloud Browser (default https://lichess.org) to work with accept team join requests and message your team.

Inputs: url.

lichess_readBInspect

Read the current Lichess page: its text and numbered buttons, links and fields.

Inputs: element.

lichess_statusBInspect

Check that your Cotlet.dev account and Lichess are ready: account, credits and Lichess connection.

Inputs: no named parameters.

lichess_typeBInspect

Type text into a numbered field on the current Lichess page.

Inputs: text (required), element (required).

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.7/5.0

    Scored across 7 tools

    Disambiguation5/5

    Each tool maps to a distinct step in a browser automation loop: guide/status for setup, open for navigation, read for page inspection, click/type for interaction, and close for teardown. There is no meaningful overlap between any pair, so an agent can select the right tool unambiguously.

    Naming Consistency4/5

    All tools share a consistent lichess_ prefix and snake_case style, which is predictable and readable. Two tools (guide, status) are nouns rather than verbs while the rest are action verbs, a minor deviation from a strict verb_noun pattern.

    Tool Count5/5

    Seven tools is a tight, well-scoped set that covers the full browser automation lifecycle for the stated Lichess workflow without filler. Each tool clearly earns its place.

    Completeness4/5

    The core loop (open, read, click, type, close) plus setup (guide/status) fully covers the 'accept team join requests and message team' purpose. Minor gaps exist—no explicit scroll, back/navigate, or wait/screenshot—but the read+click+type combination lets agents work around most of these.

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      D
      maintenance
      Enables AI agents to control the user's Chrome or Firefox browser, leveraging existing sessions for tasks requiring authentication and user handoff.
      18
      58 npm
      21
      MIT
    • A
      license
      A
      quality
      C
      maintenance
      Enables AI assistants to browse profiles and companies, manage messages and inboxes, and search jobs, posts, and people through a locally logged-in browser session. All actions run on the user's own machine.
      19
      Apache 2.0
    • A
      license
      A
      quality
      B
      maintenance
      Enables AI agents to use an existing browser session to read profiles and companies, search jobs and people, and manage messages and connection requests on LinkedIn.
      21
      Apache 2.0
    • A
      license
      B
      quality
      B
      maintenance
      Enables AI agents to browse, click, type and read on any website the user is logged in to, using encrypted local sessions with multiple named accounts per site. It behaves like a human, asks for confirmation before risky actions, and hands control back for CAPTCHAs or 2FA.
      28
      MIT