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

Your strategy against a million counted strategies

Set a strategy's returns against every strategy in the Lab (about a million rules on public portfolio and factor data, built on this machine): the nearest ones by correlation of monthly returns, how many are near-copies, and where your Sharpe ranks among all strategies covering the same months, with the best of them as the luck bar simple rules reached, and among blends of two of them (every pair, hundreds of billions). Daily returns are compounded into months; at least 24 shared months are needed.

The tool is originality_check 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
returnsyesnumber[]
datesyesstring[]
nearestnointeger

How a model reaches it

Ask in plain words. The model finds originality_check 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": "originality_check",
  "arguments": {
    "returns": "<number[]>",
    "dates": "<string[]>"
  }
}

Words that find it

Besides its name, title and description, find_tool matches a question to originality_check on these words: originality, novelty, is, my, strategy, new, nearest, strategies, correlation, duplicate, copy, crowded, factor, exposure, luck, bar, percentile, beat, million, lab.

More

Nothing here is investment advice.