AI model documentation completeness: free MCP tool
Call the AI model documentation completeness from your AI as a free MCP tool, no account: the inputs, a worked request and response, and how to add the server.
ai_model_documentation_score is one of the 20 free GRC tools from ISMS Copilot that any AI client can call over MCP: no account, no model, one call.
Scores how complete the documentation of one AI model or system is (model card, dataset datasheet, technical file), in 14 areas that follow the EU AI Act's Annex IV and Annex XI structure read with ISO/IEC 42001:2023: overall score and band, a score per section, and what to document first. A completeness aid, not a conformity assessment. Send a 0 to 4 rating for every area (0 = Not documented, 1 = Started / notes only, 2 = Partly documented, 3 = Mostly documented, 4 = Documented & reviewed); with any area missing the tool returns the open areas instead of a score. Returns general information from a fixed rule set; confirm against the official text.
This page is about calling the tool from an AI client: the request, the response and the setup. For the same tool on the web, with the background on the rules it applies, open the AI model documentation completeness page.
Inputs
The tool's input schema, as tools/list returns it, field by field.
| Field | What it is | Allowed values |
|---|---|---|
scores | One rating per area, 0 to 4: 0 = Not documented, 1 = Started / notes only, 2 = Partly documented, 3 = Mostly documented, 4 = Documented & reviewed. Every area needs a rating before a score is computed. | object (fields below) |
scores.intended-purpose | Intended purpose and use context. You have written down what the system or model is for, who is meant to use it, in what context, and the conditions, assumptions, and limits of its intended use, including reasonably foreseeable misuse to avoid. (EU AI Act Annex IV(1); Annex XI for GPAI models.) | whole number, 0 to 4 |
scores.system-description | General description, versions, and acceptable use. A reviewer can follow what the system does and how it is provided (as a product, an API, or an embedded component), which version or release this documentation covers, and any acceptable-use terms or licence attached to it. (Annex IV(1); Annex XI model description.) | whole number, 0 to 4 |
scores.data-provenance | Data sources and provenance. You record where the training, validation, and test data came from, how it was collected or obtained, and the licensing or legal basis under which you use it. (Annex IV(2); Article 10; Annex XI data information.) | whole number, 0 to 4 |
scores.data-composition | Dataset composition and preparation. You document the datasets used to build and evaluate the system: their size and main characteristics, how they were labelled, and the cleaning, filtering, or other preparation applied. (Annex IV(2); Annex XI data curation.) | whole number, 0 to 4 |
scores.data-governance | Data quality and bias examination. You document how data quality and relevance are judged, and how possible biases in the data and their effects on the people the system affects are examined and addressed. (Article 10; Annex XI bias-detection measures.) | whole number, 0 to 4 |
scores.model-design | Design choices and architecture. You document the key design decisions, the general logic and method of the system, the model type and main parameters at an appropriate level, and the rationale for the principal choices, without having to disclose trade secrets. (Annex IV(2); Annex XI architecture and parameters.) | whole number, 0 to 4 |
scores.model-performance | Performance metrics and evaluation. You report the metrics used to measure accuracy and performance, why those metrics fit this system, the results on representative data, and how the system was tested and validated. (Annex IV(2), (3), and (4); Annex XI evaluation results.) | whole number, 0 to 4 |
scores.model-limitations | Known limitations and expected accuracy. You state the system's known limitations, the conditions under which performance degrades, the level of accuracy users should expect, and any group for which it may perform differently. (Annex IV(3).) | whole number, 0 to 4 |
scores.risk-management | Risks to people and mitigations. You document the foreseeable risks the system poses to health, safety, fundamental rights, or other people, and the measures taken to identify, evaluate, and reduce them. (Annex IV(5); Article 9.) | whole number, 0 to 4 |
scores.human-oversight | Human oversight measures. You document how people can understand, monitor, intervene in, or override the system, and the technical and organisational measures built in to make that oversight effective. (Annex IV(2) and (3); Article 14.) | whole number, 0 to 4 |
scores.transparency-instructions | Information and instructions for users. You give the people deploying or using the system clear instructions: its capabilities and limitations, the conditions required for safe use, and what is expected of a human overseeing it. (Article 13.) | whole number, 0 to 4 |
scores.robustness-security | Robustness, accuracy, and cybersecurity. You document the measures that keep the system accurate and resilient against errors and adversarial conditions, and that protect the system and its data against cybersecurity threats, consistent with its intended use. (Annex IV(2); Article 15.) | whole number, 0 to 4 |
scores.lifecycle-monitoring | Versioning, change, and post-deployment monitoring. You version the documentation with the system, record the relevant changes made across its lifecycle, and have a plan to monitor how it performs once it is in use. (Annex IV(6) and (9); Article 72; ISO/IEC 42001:2023 lifecycle expectations.) | whole number, 0 to 4 |
scores.third-party-resources | Third-party components, standards, and resources. You identify the third-party tools, pre-trained or general-purpose models, and other externally-sourced components the system relies on, the standards or alternative solutions you applied, and, for larger models, the computational resources used. (Annex IV(2) and (7); Annex XI computational resources.) | whole number, 0 to 4 |
If an answer is missing
The tool never guesses an answer. A call with any field left out returns status: "needs_answers" and the open questions, each with its id, its text and its allowed values, and no result. Your AI can ask you those questions and call again.
Worked example
This request and response come from the tool itself: the catalog runs the tool's handler on this example input, and this page is generated from that catalog.
Request
A tools/call request, as your client sends it:
{
"jsonrpc": "2.0",
"id": 1,
"method": "tools/call",
"params": {
"name": "ai_model_documentation_score",
"arguments": {
"scores": {
"intended-purpose": 4,
"system-description": 3,
"data-provenance": 2,
"data-composition": 2,
"data-governance": 1,
"model-design": 3,
"model-performance": 2,
"model-limitations": 1,
"risk-management": 1,
"human-oversight": 2,
"transparency-instructions": 2,
"robustness-security": 1,
"lifecycle-monitoring": 0,
"third-party-resources": 2
}
}
}
}Response
The structuredContent of the result. The result's text content says the same in plain words, with the disclaimer.
{
"status": "result",
"overall": 1.86,
"band": "partial",
"band_label": "Partial",
"groups": [
{
"kind": "section",
"id": "purpose",
"label": "Purpose and intended use",
"score": 3.5,
"band": "documented"
},
{
"kind": "section",
"id": "data",
"label": "Data and datasets",
"score": 1.67,
"band": "partial"
},
{
"kind": "section",
"id": "model",
"label": "Model design and performance",
"score": 2,
"band": "partial"
},
{
"kind": "section",
"id": "risk",
"label": "Risk, oversight, and transparency",
"score": 1.67,
"band": "partial"
},
{
"kind": "section",
"id": "governance",
"label": "Lifecycle, security, and provenance",
"score": 1,
"band": "major-gaps"
}
],
"priorities": [
{
"id": "lifecycle-monitoring",
"title": "Versioning, change, and post-deployment monitoring",
"group": "governance",
"score": 0,
"score_label": "Not documented"
},
{
"id": "data-governance",
"title": "Data quality and bias examination",
"group": "data",
"score": 1,
"score_label": "Started / notes only"
},
{
"id": "model-limitations",
"title": "Known limitations and expected accuracy",
"group": "model",
"score": 1,
"score_label": "Started / notes only"
},
{
"id": "risk-management",
"title": "Risks to people and mitigations",
"group": "risk",
"score": 1,
"score_label": "Started / notes only"
},
{
"id": "robustness-security",
"title": "Robustness, accuracy, and cybersecurity",
"group": "governance",
"score": 1,
"score_label": "Started / notes only"
}
],
"rated": 14,
"total": 14,
"basis": [
"EU AI Act Annex IV",
"EU AI Act Annex XI",
"EU AI Act Articles 9, 10, 11, 13, 14, 15, 53 and 72",
"ISO/IEC 42001:2023"
],
"sources_checked_on": [
{
"source": "EU AI Act, Regulation (EU) 2024/1689, consolidated text of 27 July 2026 (as amended by Regulation (EU) 2026/1744)",
"checked_on": "2026-09-29"
},
{
"source": "ISO/IEC 42001:2023, Artificial intelligence management system",
"checked_on": "2026-09-25"
}
],
"disclaimer": "General information from a fixed rule set, not legal advice. Check the result against the official text and with a qualified professional before you act on it.",
"more": {
"url": "https://www.ismscopilot.com/mcp?ref=mcp-tool-ai-model-documentation-checker",
"label": "Review your own documents against this framework with ISMS Copilot"
},
"tool_page": "https://www.ismscopilot.com/resources/ai-model-documentation-checker"
}What the result rests on
Every result names its basis, as identifiers only (never standard text):
- EU AI Act Annex IV
- EU AI Act Annex XI
- EU AI Act Articles 9, 10, 11, 13, 14, 15, 53 and 72
- ISO/IEC 42001:2023
Sources the rules were checked against, with the date each was last checked:
- EU AI Act, Regulation (EU) 2024/1689, consolidated text of 27 July 2026 (as amended by Regulation (EU) 2026/1744), checked 2026-09-29
- ISO/IEC 42001:2023, Artificial intelligence management system, checked 2026-09-25
Add the server
claude mcp add --scope user --transport http ismscopilot-free-tools 'https://www.ismscopilot.com/api/public/tools/mcp'That is Claude Code. For claude.ai, Claude Desktop, ChatGPT, Cursor, Codex and any other client, see Add the server to your client. No sign-in, no token.
Limits
General information from a fixed rule set, not legal advice. Check the result against the official text and with a qualified professional before you act on it.
Your inputs are never logged or stored by our code. We count calls per tool (the tool name, whether it produced a result and when, nothing else). Our host keeps standard request logs (time, path, status, client address) for a short period. Do not send personal data: no tool needs any.
A fixed rule cannot read your own policies, registers or contracts. For a review of your own documents, connect ISMS Copilot, the GRC expert your AI calls.