canli-mcp / The desk, built into canli-mcp

Count the trials behind a result, from the signed log

How many different backtests ran through this server, counted from its signed audit log, not from memory: the number to give validate_deflated_sharpe or validate_luck_trials, which someone else can check with canli-mcp audit verify. Needs the audit log on (CANLI_POLICY).

The tool is signed_trial_count in canli-mcp 0.22.0, free and MIT-licensed, with no account and no API key. It is part of canli-mcp itself, so it is there as soon as canli-mcp is added.

Arguments

As the server declares them
ArgumentRequiredType
sincenostring
untilnostring
toolsnostring[]

How a model reaches it

Ask in plain words. The model finds signed_trial_count with canli-mcp's find_tool, reads its arguments with describe_tool if it needs to, and runs it with run_tool. The call has this shape, with your values in place of the angle brackets:

{
  "name": "signed_trial_count",
  "arguments": {}
}

Words that find it

Besides its name, title and description, find_tool matches a question to signed_trial_count on these words: number, of, trials, how, many, backtests, tried, multiple, testing, honest, count, signed, audit, proof, selection, bias, deflated, sharpe, p, hacking, data, snooping.

More

Nothing here is investment advice.