Skip to main content
Glama
README.md
# Sleeper API and MCP server

Sleeper has a public read-only API and no official MCP server. Cotlet.dev offers a hosted MCP connection to your signed-in Sleeper account with 13 tools, including six write operations for lineups, waivers, trades and league chat.

## What you can ask

> Bench starters who are ruled out in every league.

> Place a FAAB waiver bid after checking the available free agents.

> Show my matchups and scores for this week.

> Draft the weekly league recap for my chat.

> Answer open trade offers after I approve the changes.

## Connect in 2 minutes

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

### Claude Code

```bash
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:

```json
{
  "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 Sleeper 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.

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

Or with Docker:

```bash
docker build -t sleeper-mcp .
docker run -i --rm -e COTLET_API_KEY sleeper-mcp
```

Tools: `sleeper_guide`, `sleeper_status`, `sleeper_open`, `sleeper_read`, `sleeper_click`, `sleeper_type`, `sleeper_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:

```bash
pip install sleeper-mcp        # or run it once: uvx sleeper-mcp connect
sleeper-mcp connect
```

Or straight from this repo:

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

## Questions from Sleeper users

**Does Sleeper have an MCP server or an API?**

Sleeper has a public read-only API and no official MCP server. Cotlet.dev offers 13 hosted tools over MCP and REST, including writes that ask for approval.

**Which account do I connect?**

Connect your own Sleeper account in its Cotlet Cloud Browser, sign in yourself and enable it for agents.

**Which actions ask for approval?**

Lineup changes, waiver bids, trades and league chat messages require your approval before they are sent.

**Is this affiliated with Sleeper?**

No. Cotlet.dev is independent and is not affiliated with, endorsed by or sponsored by Sleeper.

## Find it on

- [Sleeper on Cotlet.dev](https://cotlet.dev/integrations/sleeper-api-mcp-server?utm_source=github&utm_medium=readme&utm_campaign=sleeper&utm_content=find)
- [PyPI package sleeper-mcp](https://pypi.org/project/sleeper-mcp/)
- [Odel](https://odel.app/modules/cotlet/sleeper)
- [mcpbeat](https://mcpbeat.com/mcp-servers/cotlet/sleeper/)
- [Step-by-step guide (PDF) on Internet Archive](https://archive.org/details/cotlet-connect-sleeper-to-ai-agents-guide)

## Also works with


- [All apps connected through Cotlet.dev](https://github.com/cotlet-connect)

---

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

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

TDQS

A3.7/5.0

Scored across 7 tools

Disambiguation5/5

Each tool maps to a distinct action: open/close session, read page, click element, type text, check status, read guide. The browser-automation verbs are unambiguous and there is essentially no overlap between click, type, and read.

Naming Consistency5/5

All seven tools share the sleeper_ prefix followed by a single clear verb (open, close, read, click, type, status, guide). The pattern is uniform and highly predictable.

Tool Count5/5

Seven tools is well-scoped for a browser automation server, with each tool earning its place across the session lifecycle. Nothing feels padded or bloated.

Completeness4/5

The set covers a coherent session lifecycle: guide, status, open, read, click, type, close. Minor gaps exist for common browser interactions like scrolling, navigation/back, or waiting on dynamic content, but core workflows are fully supported.