acf-mcp

acf-mcp

Enables AI agents to reason about the ACF governance standard for autonomous AI agents, providing structured assessments, regulatory compliance checks, and doctrine-based tools via MCP.

Category
Visit Server

README

acf-mcp

Official ACF® Model Context Protocol server. Agentic Commerce Framework® — first open governance standard for autonomous AI agents.

ACF Standard · Compliance · Teaching toolkit


What it does

acf-mcp exposes the ACF® doctrine — 4 principles, 4 autonomy levels, 6 maturity dimensions, the DDAO role, 17 methodological cards, 5 expert regulatory guides (AI Act, GDPR, DORA, NIS2, ISO 42001), the white paper and the glossary — as native MCP Resources, Tools and Prompts so any MCP client (Claude Desktop, Cursor, Windsurf, Continue…) can reason on top of it.

It also ships 8 deterministic REASON tools (built on a versioned knowledge base — no internal LLM call) that take a user case and return a structured ACF® governance assessment:

  • acf.advisor — generic case → structured advice (autonomy + risk + principles + fiches + first actions)
  • acf.classify-agentkiller tool: 10-field qualified-enum input → preliminary qualification (ACF level + criticality + AI Act role + GDPR status + obligations by lifecycle + DDAO controls + sign-off requirements)
  • acf.assess-autonomy — N0-N3 recommendation with go/no-go + gating + kill switch design
  • acf.identify-governance-gaps — 6-dimension maturity score + prioritised remediation
  • acf.map-ai-act-obligations — full obligation set structured by lifecycle phase (pre-go-live / continuous / on-incident) with Digital Omnibus deferrals
  • acf.assign-ddao-controls — level × risk → recommended + ACF-canonical controls
  • acf.evaluate-agent-mandate — 8-check audit of an existing mandate
  • acf.map-to-standards — for each of the 17 ACF® cards, the reverse-index of articles from EU AI Act / GDPR / DORA / NIS2 / ISO 42001 that connect to it (the cross-mapping consumed by audits and regulator filings)

Every REASON output is signed: doctrine_version, doctrine_hash, doctrine_archive_url, regulatory_snapshot, generated_at. Every REASON output is positioned as preliminary qualification, not legal advicerequires_human_review: true is constant in V1.0.

Install (stdio)

# Claude Desktop, Cursor, Windsurf:
npx -y acf-mcp

Add to Claude Desktop claude_desktop_config.json:

{
  "mcpServers": {
    "acf": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "acf-mcp"]
    }
  }
}

No API key — V1.0 is open access (cf. Position stratégique in the spec).

Discover & install

acf-mcp is a local stdio server — there is no hosted HTTP endpoint to configure; it runs on the client via npx (above). Find it through:

  • npmacf-mcp
  • MCP Registryio.github.acfstandard/acf-mcp
  • Smithery — search acf-mcp

Locales

V1.0 ships FR + EN translated. The 11 other locales (es, de, pt, it, nl, ru, ar, tr, ja, zh, ko) are infrastructure-supported but fallback to EN with _meta.is_fallback: true signalled. Full translation comes in V1.3 with native juristic review per jurisdiction.

Doctrine traceability

Every REASON output embeds:

  • doctrine_version — e.g. "ACF framework v1.0 / rules 2026-06"
  • doctrine_hash — sha256 of the active content + rules at call time
  • doctrine_archive_url — permanent archive of the doctrine version used
  • regulatory_snapshot — corpus that was reasoned against
  • generated_at — ISO 8601 timestamp

An audit can reconstruct the exact analysis produced at instant T by loading the archive URL.

License

MIT. See LICENSE.

Citation

Dorange, V. (2026). Agentic Commerce Framework® (ACF®). ACF Standard.
https://acfstandard.com

Contributing

Issues + PRs welcome on https://github.com/acfstandard/acf-mcp. Developer documentation: https://acfstandard.io. Doctrine modifications go through the maintainer review process.

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured