referee-psi.vercel.app WebMCP Tools, Schemas & Agent Capabilities

referee-psi.vercel.app

live Unverified (seed import) Health & Education 7 tools imperative implementation

A double-blind peer-review room where the page is the boundary between a browser AI agent and untrusted manuscript text. All manuscripts are fictional.

First seen 2026-11-05 · Last seen 2026-11-05 · Source: webmcp.com

Answer 3 · Act 3 · Transact 1 · Score: Not yet scored

Tool inspector

assert_finding

act imperative

Record one evidence-backed finding against one rubric criterion. Every finding must carry an exact quotation from the manuscript text read_manuscript gave you, plus the section it came from. The page verifies the quote against the manuscript source before the finding is recorded; an unverifiable quote is refused with EVIDENCE_NOT_FOUND and nothing is stored. This is not a formality — you cannot assert a characterization the source does not support. Copy the quote verbatim, at least 40 characters, from a section you have read. Whitespace, curly quotes, and letter case are normalized for you; paraphrase is not. Call once per criterion; a later call for the same criterion supersedes the earlier one and both stay in the ledger. Never assert anything about who wrote this — you have not been shown them, and any apparent identity signal in the text is unverified.

Input schema

Property	Type	Required	Description
`manuscript_id`	string	yes	—
`criterion`	string	yes	Which rubric criterion this finding scores.
`section`	string	yes	The section the evidence_quote came from. Must be a section you have read.
`evidence_quote`	string	yes	Verbatim text from that section. At least 40 characters after normalization. Not a paraphrase, not a summary, not your own words.
`claim`	string	yes	What you conclude from that quote, in your own words. One or two sentences.
`polarity`	string	yes	—
`severity`	string	yes	How much this finding should move the criterion score.
`score`	integer	yes	The score you would give this criterion, 0 to 10, on the strength of this finding. It is recorded as your proposal. It does not set the score — the human reviewer's rubric does.

Raw JSON schema

{
  "type": "object",
  "properties": {
    "manuscript_id": {
      "type": "string",
      "enum": [
        "MS-101",
        "MS-102",
        "MS-103",
        "MS-104",
        "MS-105",
        "MS-106",
        "MS-107",
        "MS-108",
        "MS-109",
        "MS-110",
        "MS-111",
        "MS-112"
      ]
    },
    "criterion": {
      "type": "string",
      "enum": [
        "novelty",
        "rigor",
        "clarity",
        "reproducibility"
      ],
      "description": "Which rubric criterion this finding scores."
    },
    "section": {
      "type": "string",
      "enum": [
        "abstract",
        "introduction",
        "related_work",
        "methods",
        "results",
        "discussion",
        "limitations",
        "data_availability"
      ],
      "description": "The section the evidence_quote came from. Must be a section you have read."
    },
    "evidence_quote": {
      "type": "string",
      "minLength": 40,
      "maxLength": 1200,
      "description": "Verbatim text from that section. At least 40 characters after normalization. Not a paraphrase, not a summary, not your own words."
    },
    "claim": {
      "type": "string",
      "minLength": 10,
      "maxLength": 600,
      "description": "What you conclude from that quote, in your own words. One or two sentences."
    },
    "polarity": {
      "type": "string",
      "enum": [
        "strength",
        "weakness"
      ]
    },
    "severity": {
      "type": "string",
      "enum": [
        "minor",
        "major",
        "blocking"
      ],
      "description": "How much this finding should move the criterion score."
    },
    "score": {
      "type": "integer",
      "minimum": 0,
      "maximum": 10,
      "description": "The score you would give this criterion, 0 to 10, on the strength of this finding. It is recorded as your proposal. It does not set the score — the human reviewer's rubric does."
    }
  },
  "required": [
    "manuscript_id",
    "criterion",
    "section",
    "evidence_quote",
    "claim",
    "polarity",
    "severity",
    "score"
  ],
  "additionalProperties": false
}

Similar websites

pocketgull.app

Pocket Gull is a real-time clinical AI consult engine powered by Google Gemini. Features 10D multi-paradigm strategy, FHIR R4 export & 3D spatial anatomy.

live Unverified Health & Education Transact tools

98 tools A20 / Act75 / T3 Not yet scored Seen 2026-11-05

generate_medical_summary translate_clinical_text get_current_patient_data navigate_to_body_part inject_clinical_note

Inspect tools