Log inSubmit server
Canli Validation MCP logo

Canli Validation MCP

Validates backtests with deflated Sharpe, CSCV, track-record, haircut, and luck-equivalent-trial checks.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Canli Validation MCP

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "canli-validation-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "canli-validation-mcp"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Automated check passed— started and listed 19 tools correctly (Today).

Overview

canli-validation-mcp connects AI assistants to Canli Capital's backtest-validation service through MCP. It provides checks including deflated Sharpe, CSCV overfitting probability, minimum track record, haircut Sharpe, and luck-equivalent trials. Each result includes a signed receipt that can be independently recomputed. Reach for it when assessing whether a backtest's reported performance survives common selection and overfitting checks; a result is not an admission decision or forecast.

Use cases

  • Validate a reported Sharpe ratio against the number of trials run
  • Estimate the probability that a backtest is overfit
  • Check whether a strategy has enough track record
  • Apply a conservative haircut to backtest performance
  • Create independently recomputable validation evidence

Key features

  • Deflated Sharpe validation
  • CSCV overfitting-probability checks
  • Minimum-track-record analysis
  • Haircut Sharpe calculations
  • Luck-equivalent-trial analysis
  • Signed, recomputable receipts

Tools19

What Canli Validation MCP lets your agent do. Expand a tool for its parameters.

Verified live~6.4k tokens of context

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 3d ago

Recent health check succeeded.

6/6 checks healthy over the last 45d

What it can do

From the 19 tools this server returned when we ran it; 19 declare their own safety hints.

Makes network requests · 13Read-only tools · 19
Network at startup3 connections
Files written at startupNone
Install scriptsNone
npm provenanceNot published
Package age50 days
Maintainers1

Observed 2026-10-27 in an isolated sandbox. A description of reach, not a security audit.

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 11d ago via OSV.dev · canli-validation-mcp (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars5Stargazers on the source repository.
npm downloads4.4kPackage downloads in the last 30 days.
Last commit3d agoMost recent push to the default branch.
Availability100%Our rolling endpoint + install checks that succeeded.
Install checkPassedOur sandbox started it and listed its tools.
Tools exposed19Callable tools this server registers over MCP.

How Canli Validation MCP works

What canli-validation-mcp does

canli-validation-mcp exposes Canli Capital's backtest validation capabilities to MCP-compatible AI assistants. Its purpose is to examine reported backtest performance using several statistical and research-process checks rather than treating a single headline metric as sufficient evidence.

The available validation methods include deflated Sharpe, Combinatorially Symmetric Cross-Validation (CSCV) overfitting probability, minimum track record, haircut Sharpe, and luck-equivalent trials. These checks address different concerns around selection, repeated experimentation, track-record length, and the reliability of an observed Sharpe ratio.

Results come with signed receipts. The receipt is intended to make the calculation auditable: an independent party can recompute the result rather than relying only on the assistant's response.

How it works

The server runs as an MCP process that an assistant launches locally. The repository's quick-start command uses npx -y canli-validation-mcp, so the package can be started without a separate application install. Claude Code is shown as one launch target, with setup instructions also identified for Cursor, VS Code, and Claude Desktop.

An assistant can use the server when reviewing a strategy or backtest and request the relevant validation calculation. The resulting check and its signed receipt provide a machine-readable basis for further review. Canli Capital also exposes a separate free HTTP API for validation, but that API is distinct from using the MCP package.

Setup and configuration

Install and register the package with an MCP client using the command below:

Terminal
claude mcp add canli -- npx -y canli-validation-mcp

The command is shown for Claude Code. Client-specific configuration is available for Cursor, VS Code, and Claude Desktop through the project's MCP setup materials. The MCP launch example does not require an environment variable or API key.

A separate API workflow can create a key at https://canlicapital.com/api/v1/keys and send validation requests to the public API. That key-based route should not be confused with the local MCP registration command.

Tools and capabilities

  • Calculate a deflated Sharpe ratio that accounts for the number of trials.
  • Estimate backtest overfitting probability with CSCV.
  • Evaluate whether a strategy has a sufficient minimum track record.
  • Apply a haircut to a reported Sharpe ratio.
  • Calculate luck-equivalent trials.
  • Return signed receipts that support independent recomputation of results.

Limitations and notes

Validation output is not an admission decision and should not be treated as a forecast. The checks are intended to assess the credibility of backtest evidence; they do not turn a validated result into a guarantee about future performance.

The repository distinguishes this validation server from two other MCP packages in the same project: one for point-in-time SEC fundamentals and another for Canli Capital's research record. Those capabilities are provided by separate packages rather than by canli-validation-mcp.

The project publishes its validation service alongside open research and reproducibility artifacts. The signed receipt model is useful for review workflows where the calculation needs to be checked outside the assistant that requested it.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Canli Validation MCP

canli-validation-mcp is an MCP server that connects AI assistants to Canli Capital's backtest-validation service and exposes checks for deflated Sharpe, CSCV overfitting probability, minimum track record, haircut Sharpe, and luck-equivalent trials. It returns validation results with signed receipts that can be recomputed independently.

Install it through an MCP client with `npx -y canli-validation-mcp`; for Claude Code, run `claude mcp add canli -- npx -y canli-validation-mcp`. The project also identifies setup paths for Cursor, VS Code, and Claude Desktop.

It provides deflated Sharpe, CSCV overfitting probability, minimum track record, haircut Sharpe, and luck-equivalent-trial checks. Each validation result includes a signed receipt for independent recomputation.

Yes, Cursor is listed as a supported setup target. The server runs as an MCP process, and the project includes Cursor configuration guidance alongside instructions for Claude Code, VS Code, and Claude Desktop.

No, its verdict is not a forecast. The server evaluates backtest evidence with statistical validation checks, but its output should not be treated as a guarantee of future performance or as an admission decision.

Related MCP servers

All in Finance & Fintech All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewCanli Validation MCP AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/canli-validation-mcp?style=directory)](https://allmcps.com/mcp/canli-validation-mcp)
HTML Embed
<a href="https://allmcps.com/mcp/canli-validation-mcp"><img src="https://allmcps.com/api/badge/canli-validation-mcp?style=directory" alt="Canli Validation MCP on AllMCPs" /></a>