Metrc

A Model Context Protocol (MCP) server that provides AI agents with searchable access to METRC cannabis compliance documentation across all legal states.

🚀 Live Demo: https://www.metrc.ai

What This Does

Enables AI agents (Claude Desktop, Cursor, etc.) to search and reference 7,282 sections of METRC documentation across 17 states in real-time during conversations.

Quick Start

For Claude Desktop Users

1. Visit https://www.metrc.ai
2. Download the connector file
3. Drag & drop it into Claude Desktop
4. Ask: "What are the package labeling requirements in California?"

For Developers

# Clone and setup
git clone https://github.com/your-username/metrc-mcp-server.git
cd metrc-mcp-server
bundle install

# Import documentation (if you have the source data)
bin/import /path/to/metrc_from_printable

# Start MCP server (stdio mode)
bin/mcp_server

# Or start HTTP server
bin/simple_mcp_server

Available Tools

• search_metrc_docs - Full-text search across all documentation
• get_metrc_endpoint - Find API endpoint documentation
• list_states - Get all available states

States Covered

CA, CO, NV, OR, AK, MA, MI, IL, MD, OH, OK, LA, MN, MO, DC, VA, WV

Architecture

• SQLite FTS for fast full-text search
• MCP Protocol for AI agent integration
• HTTP/JSON-RPC transport for web deployment
• Ruby/Sinatra lightweight server

Deployment

Includes deployment scripts for EC2/VPS hosting:

./deploy.sh  # Deploy to production server

API Example

curl -X POST https://www.metrc.ai/mcp \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "id": 1,
    "method": "tools/call",
    "params": {
      "name": "search_metrc_docs",
      "arguments": {"query": "packages"}
    }
  }'