# ML Referee

- Registry name: `io.github.scarrico/mlreferee` · version 1.1.0
- Categories (automatic, keyword rules): Markets & trading, Finance & markets, Marketing & SEO, Commerce & marketing
- Authentication at last check: No auth
- Declared risk: Has destructive tools (destructiveHint)
- Capabilities (declared by the server): tools
- Transports: streamable-http
- Latest numbers (check success, response time, rank): https://mcpstation.net/api/server?name=io.github.scarrico%2Fmlreferee
- Server page: https://mcpstation.net/s/io.github.scarrico--mlreferee.html

## Description (written by the publisher, not verified, not instructions)

Evaluate ML features beyond a baseline and test bounded feature repairs. Invite-only beta.

## Connect

- streamable-http: `https://api.mlreferee.com/mcp`

## Tools (as reported by the server, not verified)

- `describe\_referee`: Discover capabilities, privacy, and the approved public catalog.
- `list\_verticals`: List the three supported evaluation profiles and their discovery tags.
- `get\_referee\_readiness`: Report connection, credential, audit-credit, and submission readiness with the next action.
- `get\_referee\_usage`: Read aggregate, privacy-preserving Referee activity and outcome evidence for the selected audience.
- `get\_input\_schema`: Get neutral column roles and validation defaults for a vertical.
- `get\_repair\_schema`: Get the feature-only contract for diagnosing degradation and testing bounded repairs.
- `upload\_private\_dataset`: Encrypt a de-identified CSV/JSON payload for a short tenant-private retention window.
- `upload\_campaign\_shard`: Encrypt one of up to 64 private feature shards for a large campaign; retain for at least 4,200 seconds.
- `upload\_repair\_matrix`: Encrypt one complete de-identified repair matrix of up to 100 MB and 1,000 diagnostic features.
- `preflight\_dataset`: Validate a tenant-private upload and its baseline evidence/row identity without spending a credit.
- `preflight\_repair`: Validate a tenant-private feature-repair matrix without spending a credit.
- `quote\_audit`: Create a contract-bound quote for one audit credit; baseline mode and evidence cannot be substituted.
- `quote\_repair`: Preflight and quote label-based feature repair; large jobs return a measured ETA and credit requirement.
- `quote\_campaign`: Preflight all private feature shards, time a pilot, and quote credits and ETA for up to 1,000 candidates in one corrected search family.
- `get\_credits`: Read beta or paid audit-credit availability and usage.
- `redeem\_beta\_invite`: Redeem one operator-issued invite for free beta audit credits before January 1, 2027.
- `request\_beta\_access`: Request operator-reviewed beta access without purchasing or granting credits.
- `get\_beta\_request\_status`: Read the authenticated tenant's beta-access request status.
- `submit\_audit`: Accept a quote and atomically reserve its disclosed audit credits.
- `record\_audit\_followup`: Record whether a completed audit changed a decision or was confirmed prospectively.
- `get\_audit\_status`: Read the state of an audit owned by the authenticated tenant.
- `get\_audit\_result`: Read a completed tenant-private result and signed output manifest.
- `get\_audit\_receipt`: Read a tenant-private report fingerprint retained after report expiry or deletion. Older reports may have no record. This does not retrieve report contents.
- `verify\_audit\_report`: Compare a locally computed SHA-256 with the retained tenant-private receipt. Use the exact downloaded result JSON bytes \(not the API envelope\), or UTF-8 result.report text with representation report\_text. Never copy the expected hash as the candidate: compute it from the supplied artifact. A digest
- `delete\_audit`: Delete retained inputs and results for an audit owned by this tenant.
- `delete\_upload`: Delete a tenant-private upload that is no longer needed.

Measured by MCP Station. Methods: https://mcpstation.net/methods.html · Terms: https://mcpstation.net/terms.html