About governmyai
governmyai is an MCP server for querying AI governance obligations across regulatory and standards frameworks. It is intended for teams assessing AI systems; its results describe obligations, not compliance verdicts.
Key features
- 01Query obligations across seven frameworks
- 02Classify EU AI Act risk tiers
- 03Retrieve obligation details and evidence requirements
- 04Map obligations across frameworks
- 05Mark when human review is required
Use cases
- 01Identify obligations for a hiring AI deployed in the EU
- 02Review evidence needed for specified EU AI Act articles
- 03Compare obligations for a credit-scoring system
- 04Find cross-framework mappings for governance controls
MCP support
Exposes six tools to MCP clients for querying obligations, classifying EU AI Act risk tiers, retrieving rules and evidence requirements, and finding cross-references. Responses include a notice that results are not compliance verdicts.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
npx -y @governmyai/mcp-serverGOVERNMY_API_KEYSecretYour GovernMy.ai API key. Get a free one at https://governmy.ai/developersGOVERNMY_API_URLOverride the API base URL. Defaults to https://api.governmy.aiLocal servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
Tools
6Saved tool list · updated 2w ago
| Parameter | Type | Required | Description |
|---|---|---|---|
| role | string | No | Role under the relevant framework. EU AI Act: provider, deployer, importer, distributor. ISO 42001: api_consumer, fine_tuner, model_builder, reseller. |
| industry | string | No | Industry the AI system operates in. |
| riskTier | enum | No | EU AI Act risk tier. Omit if unknown — use classify_risk_tier first if you need to determine it.unacceptable | high | limited | minimal |
| useCases | array | No | Use case tokens (e.g. "resume_screening", "chatbot", "credit_decision"). |
| dataTypes | array | No | |
| frameworks | array | No | Restrict the query to these frameworks. Omit to query all. |
| jurisdiction | array | No | ISO country codes or region tags (e.g. ["EU", "US-CO"]). |
| lifecyclePhase | enum | No | design | training | validation | testing | deployment | monitoring | retirement |
| annexIiiCategory | array | No | EU AI Act Annex III high-risk category ids (e.g. "employment", "essential_services", "biometrics"). |
Details
From the project README@governmyai/mcp-server
GovernMy.ai MCP server — query AI governance obligations from Claude Desktop, Cursor, or any MCP-compatible client.
Covers EU AI Act, ISO 42001, Colorado AI Act, NIST AI RMF, HIPAA (AI provisions), SOX (AI provisions), and FTC AI guidance.
What this does
This MCP server gives an AI assistant (like Claude) the ability to query GovernMy's rules engine directly while you're working. You can ask questions like:
- "What EU AI Act obligations apply if I build a hiring AI for the EU?"
- "Which frameworks does this cover?"
- "What evidence do I need for EU AI Act Article 10 data governance?"
- "Does my ISO 42001 Clause 6 work also cover any NIST AI RMF functions?"
...and get structured regulatory obligations back as part of the conversation.
Important: this tool returns obligations, not verdicts. It never tells you "you're compliant" or "you're not compliant" — those are determinations only a human reviewer can make. Every response includes a humanReviewRequired flag where applicable.
Copy-paste into Claude
After installing + restarting, try one of these:
- "What are the mandatory-review obligations for a high-risk hiring AI deployed in the EU?"
- "Show me the evidence requirements for EU AI Act Articles 11 and 14."
- "Compare EU AI Act and Colorado AI Act obligations for a credit-scoring AI."
- "Does my HR chatbot trigger any regulatory obligations? It does resume screening."
- "What ISO 42001 Annex A controls do I need for a generative AI system?"
- "Which NIST AI RMF dimensions does EU AI Act Art. 10 data governance also satisfy?"
- "I'm building a medical imaging AI — what obligations apply under HIPAA AI provisions?"
- "List every obligation flagged as
cannotBeAutoSatisfiedacross all frameworks."
Install + configure (Claude Desktop)
- Register for an API key at https://governmy.ai/developers (free tier available)
- Open your Claude Desktop config file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
- macOS:
- Add the GovernMy server to your
mcpServersobject:
{
"mcpServers": {
"governmy": {
"command": "npx",
"args": ["-y", "@governmyai/mcp-server"],
"env": {
"GOVERNMY_API_KEY": "ooa_live_your_key_here"
}
}
}
}
- Restart Claude Desktop. Ask Claude: "What governance frameworks does GovernMy cover?" — it should call the
list_frameworkstool.
Install + configure (Cursor)
Add the server to your Cursor MCP settings (Settings → MCP Servers):
- Command:
npx - Args:
-y @governmyai/mcp-server - Environment:
GOVERNMY_API_KEY=ooa_live_your_key_here
Configure (other MCP clients)
Any client that speaks stdio MCP can run this server. Set GOVERNMY_API_KEY in the client's env for the server process.
Environment variables
| Variable | Required | Default | Purpose |
|---|---|---|---|
GOVERNMY_API_KEY | yes | — | Your GovernMy.ai API key |
GOVERNMY_API_URL | no | https://api.governmy.ai | Override API base URL (for local dev) |
Tools exposed
get_obligations— query obligations that apply to an AI system given context (risk tier, role, industry, etc.)classify_risk_tier— classify an AI system's EU AI Act risk tier (unacceptable / high / limited / minimal)list_frameworks— list the frameworks the engine covers with metadataget_rule— fetch the full payload for a single obligation by idget_evidence_requirements— fetch required evidence artifacts for a set of obligationsget_cross_references— fetch cross-framework mappings for an obligation
Every tool response includes a notAVerdict notice reminding the AI assistant that obligations are not compliance verdicts.
What this is not
- Not legal advice. The rules engine is a research tool + workflow accelerator. Compliance decisions require qualified human review.
- Not a replacement for a compliance professional. If your AI system is materially exposed to any of the covered frameworks, pair this with a real compliance program.
- Not a verdict engine. We deliberately do not return "compliant: true" or "allowed: true". This is both an ethical line and a legal one (EU AI Act Art. 14).
Support
- Docs: https://docs.governmy.ai/mcp
- API reference: https://docs.governmy.ai/api
- Issues: https://github.com/governmyai/mcp-server/issues
- Email: [email protected]
License
MIT
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.