Arcology Knowledge Node — MCP Server

Read-only MCP server for AI agents to query the Arcology One engineering knowledge base.

Tools

Tool	Purpose
read_node(domain, slug)	Retrieve full knowledge entry
search_knowledge(query, ...)	Full-text search with filters
list_domains()	All 8 domains with stats
get_open_questions(domain?, limit?)	Unanswered engineering questions
get_entry_parameters(domain?, parameter_name?)	Quantitative parameters for consistency checking
get_domain_stats()	Aggregate platform statistics

Local Development

# Install dependencies
pip install -e .

# Run with stdio (for Claude Desktop)
python server.py

# Run with Streamable HTTP (for remote access)
MCP_TRANSPORT=streamable-http python server.py

# Point to local Next.js dev server
INDEX_URL=http://localhost:3000/content-index.json python server.py

Claude Desktop Configuration

Add to ~/.claude/claude_desktop_config.json:

{
  "mcpServers": {
    "arcology": {
      "command": "python",
      "args": ["C:/Users/blues/Claude Code Projects/arcology-node/mcp/server.py"],
      "env": {
        "INDEX_URL": "https://lifewithai.ai/content-index.json"
      }
    }
  }
}

Deploy to Fly.io

cd mcp/
fly auth login
fly launch --name arcology-mcp --region dfw
fly deploy

Then point DNS: mcp.lifewithai.ai → Fly.io app.

Architecture

content-index.json (Netlify) → [HTTP fetch, cached 5 min] → MCP server (Fly.io) → agents via Streamable HTTP (/mcp)