Ask your AI about any tout.

ToutCheck’s verified records aren’t locked inside this website. Through the Model Context Protocol (MCP), any compatible AI client can pull the same evidence the site publishes — timestamped picks, settled outcomes, ROI, and closing-line value — and reason over it directly.

is_tout_legit

The evidence behind “is this tout legit?”: tracked status, badge status, sample size, W-L-P record, ROI, closing-line value, and winner announcements that match no logged pick.

alias — the tout’s public handle

get_tout_record

One alias’s full verified record: record, win %, net units, ROI, CLV, free vs premium split, and the sports covered.

alias — the tout’s public handle

get_leaderboard

The full public leaderboard: every tracked alias with its verified record, most graded picks first.

no parameters

Connect in two steps

1. Get the open engine (Python 3.10+, zero dependencies — it runs on the standard library alone):

git clone https://github.com/unlimitedappsllc/toutcheck.git

2. Add ToutCheck to your MCP client’s configuration:

{
  "mcpServers": {
    "toutcheck": {
      "command": "python3",
      "args": ["-m", "engine.mcp_server", "--db", "/path/to/ledger.db"],
      "cwd": "/path/to/toutcheck"
    }
  }
}

Then ask your model things like: “Using ToutCheck, what is this tout’s verified record — and does their closing-line value back it up?”

Evidence, not verdicts. ToutCheck publishes what a tout posted (timestamped before game time) and what actually happened — it never labels anyone a scam or a sharp. Your model reads the same numbers a human reader would and draws its own conclusions. Records appear as the public ledger fills; a hosted endpoint is on the roadmap.