{
  "markdown": "# Hokmah MCP Server\n\n**AI Agent with Architectural Memory — MCP Server**\n\nGives any AI coding agent persistent understanding of codebases via TransitionGraph, IdeaGraph, and WorldModel. Analyze impact, generate tests, write code — all from the graph.\n\n<!-- mcp-name: io.github.davidangularme/hokmah -->\n\n## Quick Start (30 seconds)\n\nAdd to your editor's MCP config (Cursor, Claude Code, VS Code, Windsurf, Cline, JetBrains):\n\n```json\n{\n  \"mcpServers\": {\n    \"hokmah\": {\n      \"type\": \"streamable-http\",\n      \"url\": \"https://hokmah.dev/mcp\"\n    }\n  }\n}\n```\n\nThen ask your agent: *\"analyze the impact of refactoring the auth module in github.com/owner/repo\"*\n\n## Available Tools\n\n| Tool | Tier | Description |\n|------|------|-------------|\n| `hokmah_analyze` | **FREE** | Impact analysis, risk score, affected files, architectural invariants |\n| `hokmah_connect_project` | **FREE** | Connect a GitHub repo, build the architectural graph |\n| `hokmah_connect_mcp` | **FREE** | Connect an external MCP server for orchestration |\n| `hokmah_generate_tests` | **PRO** | Test generation from the graph (40x fewer tokens) |\n| `hokmah_generate_code` | **PRO** | Code generation with architectural memory |\n\n## How It Works\n\nHokmah builds a persistent architectural graph from your codebase:\n\n- **TransitionGraph** — Markov model of code changes (which files change together)\n- **IdeaGraph** — 16 relation types between concepts\n- **WorldModel** — File tree, dependencies, symbols\n\nWhen you ask \"what's the impact of changing X?\", Hokmah traverses the graph instead of sending your entire codebase to an LLM. That's why `analyze` is free (zero LLM tokens) and `generate` uses 40x fewer tokens.\n\n## Pricing\n\n- **Free** — `hokmah_analyze` + `hokmah_connect_project` + `hokmah_connect_mcp` (unlimited)\n- **Pro** — `hokmah_generate_tests` + `hokmah_generate_code` (BYOK — bring your own LLM key)\n\nGet a Pro key at [hokmah.dev](https://hokmah.dev).\n\n## Editor Setup\n\n- **Cursor** — Settings → MCP → Add server → paste config\n- **Claude Desktop** — `~/Library/Application Support/Claude/claude_desktop_config.json`\n- **Claude Code** — `claude mcp add hokmah --transport streamable-http --url https://hokmah.dev/mcp`\n- **VS Code (Copilot)** — `.vscode/mcp.json` in project root\n- **Windsurf** — `~/.windsurf/mcp.json`\n- **Cline** — Settings → MCP Servers → Add\n- **JetBrains** — Settings → Tools → AI Assistant → MCP Servers\n\n## Self-Hosting\n\nThe hosted server at `https://hokmah.dev/mcp` is the recommended way to use Hokmah. To run the server yourself against your own Hokmah backend:\n\n```bash\npip install -r requirements.txt\ncp pro_keys.example.json pro_keys.json  # edit with your real PRO keys\nHOKMAH_API_BASE=http://localhost:8000 python mcp_server.py\n```\n\nEnvironment variables:\n\n- `HOKMAH_API_BASE` — upstream Hokmah API (default `http://localhost:8000`)\n- `HOKMAH_MCP_PORT` — port to listen on (default `8001`)\n- `HOKMAH_PRO_KEYS` — path to the PRO keys JSON file (default `/home/vpm/mcp-server/pro_keys.json`)\n\nA reference `systemd` unit is provided in [`hokmah-mcp.service`](hokmah-mcp.service).\n\n## Built by\n\n[Catalyst AI Research](https://catalystais.com) · Haifa, Israel\n\n## License\n\nMIT\n",
  "bytes": 3184,
  "sha": "86070f65ccce0292d3b6bacbbc2afa01efbf40cfea0f4e808dcbc74ee7ac6e70",
  "repo_slug": "davidangularme/hokmah-mcp-server",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_davidangularme_hokmah_7dfeb78d/readme"
}