Arcology Knowledge Node — MCP Server

1# Arcology Knowledge Node — MCP Server
2 
3Read-only MCP server for AI agents to query the Arcology One engineering knowledge base.
4 
5## Tools
6 
7| Tool | Purpose |
8|------|---------|
9| `read_node(domain, slug)` | Retrieve full knowledge entry |
10| `search_knowledge(query, ...)` | Full-text search with filters |
11| `list_domains()` | All 8 domains with stats |
12| `get_open_questions(domain?, limit?)` | Unanswered engineering questions |
13| `get_entry_parameters(domain?, parameter_name?)` | Quantitative parameters for consistency checking |
14| `get_domain_stats()` | Aggregate platform statistics |
15 
16## Local Development
17 
18```bash
19# Install dependencies
20pip install -e .
21 
22# Run with stdio (for Claude Desktop)
23python server.py
24 
25# Run with Streamable HTTP (for remote access)
26MCP_TRANSPORT=streamable-http python server.py
27 
28# Point to local Next.js dev server
29INDEX_URL=http://localhost:3000/content-index.json python server.py
30```
31 
32## Claude Desktop Configuration
33 
34Add to `~/.claude/claude_desktop_config.json`:
35 
36```json
37{
38  "mcpServers": {
39    "arcology": {
40      "command": "python",
41      "args": ["C:/Users/blues/Claude Code Projects/arcology-node/mcp/server.py"],
42      "env": {
43        "INDEX_URL": "https://lifewithai.ai/content-index.json"
44      }
45    }
46  }
47}
48```
49 
50## Deploy to Fly.io
51 
52```bash
53cd mcp/
54fly auth login
55fly launch --name arcology-mcp --region dfw
56fly deploy
57```
58 
59Then point DNS: `mcp.lifewithai.ai` → Fly.io app.
60 
61## Architecture
62 
63```
64content-index.json (Netlify) → [HTTP fetch, cached 5 min] → MCP server (Fly.io) → agents via Streamable HTTP (/mcp)
65```
66