Model Context Protocol

DAMM inside Claude and Google Antigravity

The DAMM MCP server brings 4-pillar decision analysis straight into your AI assistant. Ask about a decision and get scores, a Health Score and a structured verdict — without leaving the chat.

Install in 1 minute API documentation

Setup

Three steps: get an API key, configure your MCP client, test.

1 · API key — create a key in the developer dashboard and use it as the value of DAMM_API_KEY. Docs.

2 · Configure your client

Claude Desktop

Paste the JSON block below into the config file, then restart Claude.

macOS:   ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

Claude Code

A single command from the terminal (with your key):

claude mcp add damm -e DAMM_API_KEY=damm_sk_... -- npx -y damm-decision-engine

Antigravity & other clients

In Antigravity: agent panel → "…" menu → Manage MCP Servers → View raw config (mcp_config.json). Paste the same JSON block. Cursor and Windsurf use the same format.

MCP config (Claude Desktop, Antigravity, Cursor…)

{
  "mcpServers": {
    "damm": {
      "command": "npx",
      "args": ["-y", "damm-decision-engine"],
      "env": { "DAMM_API_KEY": "damm_sk_..." }
    }
  }
}

DAMM_API_KEY is required: every analysis runs on the DAMM decision engine. Get a key at dammframework.com.

Prefer self-hosting? Build the package (npm install && npm run build) and use "command": "node" with the absolute path to dist/index.js.

3 · Test

Ask your assistant:

“Use DAMM to analyze: I'm considering leaving my job for a startup, I have 12 months of savings and a family.”

It should return the 4 pillars with scores, the Health Score and the EU AI Act Art. 50 note — and the decision appears in your AI registry.

Requires a DAMM API key

What you get

Every analysis is produced by the DAMM decision engine — not by your assistant. It needs an API key with an active plan.

Verdict from the DAMM decision engine: 4 pillars (0-10) and a Health Score

Decisions saved to your account

EU AI Act (Art. 50) disclosure on every verdict

Immutable entry in the AI decision registry (audit trail)

Learn more about the AI decision registry.

Frequently asked questions

What is an MCP server?

The Model Context Protocol (MCP) is an open standard that lets AI assistants use external tools. An MCP server exposes a capability — here, DAMM decision analysis — to any compatible client, such as Claude or Google Antigravity.

Does it work with Claude and Antigravity?

Yes. Being a standard, the same server runs on Claude Desktop, Claude Code, Google Antigravity, Cursor, Windsurf and any MCP client. Only where you paste the config differs.

Do I need an API key?

Yes. The server requires an API key with an active plan: every analysis is produced by the DAMM decision engine. Without a key, the tool replies with instructions to get one — it does not run offline analysis.

What does the analysis return?

A structured verdict: a 0-10 score with confidence for each of the 4 pillars, a Health Score (0-100), a summary of recommended actions, an EU AI Act (Art. 50) disclosure and an immutable entry in your account’s AI decision registry.

Bring DAMM into your assistant

Get an API key, paste the config into your client, and every decision leaves a traceable verdict.

Request an API key API documentation

Claude, Claude Code and Google Antigravity are products of their respective owners; the DAMM MCP server is an independent client that integrates via the Model Context Protocol. Pro-mode verdicts are generated by an AI system: human oversight is recommended before acting.