Archon Specs MCP Server
AI-Native Architecture Compiler

Supercharge your AI coding assistants (Claude Desktop, Cursor, Gemini) with direct access to the Archon materialization engine. Compile intent into validated DesignSpecs and scaffold production-grade NestJS backends instantly.

Explore MCP Capabilities CLI & Client Setup

Deterministic Materialization via MCP

The Archon Specs MCP Server equips your AI assistant with specialized tools to enforce architectural guardrails, manage domain shards, and verify production readiness autonomously.

1. archon_generate

Scaffold enterprise-grade NestJS backends from validated DesignSpecs. Automatically wires controllers, services, and dependency injection.

2. archon_validate

Perform rigorous schema and semantic validation against `designspec.schema.json` to eliminate AI hallucinations before code generation.

3. archon_create_domain_shard

Structure backend modules by domain. Elicit requirements and maintain topological dependency graphs (`dependsOn`) across microservices.

4. normalizeSpec

Harden and normalize raw architectural specifications into immutable contracts, ensuring clean auto-pathing and routing configuration.

5. archon_proof

Launch automated Docker containers (Postgres, Redis) and generate live Swagger OpenAPI endpoints for immediate verification.

6. archon_lineage

Track manual developer code edits and establish strict lineage guardrails to ensure custom business logic is never overwritten.

7. archon_library

Pull curated, production-tested prompts, tools, and materialization specs directly from the official Archon Agent Library.

8. stream_telemetry

Monitor real-time compilation progress and background worker pool execution via Server-Sent Events (SSE) streaming.

Governed Token Authentication

The Archon Specs MCP Server utilizes a secure, dashboard-linked authentication workflow to protect your project specs and API keys.

  1. Generate your secure API token in the Archon dashboard.
  2. Run `npx archonspecs -token` in your terminal to authenticate.
  3. Tokens are encrypted and stored persistently in your local configuration.
  4. Seamlessly propagate `planId` and `apiKeyId` across all MCP tool calls.

Claude & Cursor Configuration

Integrating the Archon Specs MCP Server into your AI assistant takes less than a minute. Add the following block to your MCP settings:

{ "mcpServers": { "archonspecs": { "command": "npx", "args": ["-y", "archonspecs", "mcp"] } } }

Paste this directly into your `claude_desktop_config.json` or Cursor MCP configuration panel to unlock autonomous system architecture capabilities.

Ready to give your AI architectural superpowers?

Connect the Archon Specs MCP Server today and let your AI assistant compile, validate, and verify enterprise backends deterministically.


Launch Archon Specs Hub →