{
  "markdown": "# Bolor Brain MCP\n\n**Pure intelligence for Claude Code.** Reasoning, memory, and learning -- nothing else.\n\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)\n[![Python 3.11+](https://img.shields.io/badge/python-3.11+-blue.svg)](https://www.python.org/downloads/)\n[![MCP Compatible](https://img.shields.io/badge/MCP-Compatible-green.svg)](https://modelcontextprotocol.io/)\n\n---\n\n## What Is This?\n\nBolor Brain is an MCP server that gives Claude Code a **brain**: structured reasoning, persistent memory, and learning from experience.\n\n```\nUser --> Claude Code (Gateway + Executor)\n              |\n         +----+----+\n    Bolor Brain   NSAF\n      (MCP)       (MCP)\n      THINK       EVOLVE\n```\n\n- **Claude Code** = Gateway + tool execution + sessions + permissions\n- **Bolor Brain** = Reasoning engines + memory + learning + persistence\n- **NSAF** = Strategy evolution + self-improvement (separate MCP server)\n\nBolor Brain does NOT execute anything. No file ops, no scheduling, no autonomous loop. Claude Code already does all of that. Bolor Brain only thinks.\n\n---\n\n## Quick Start\n\n### 1. Install\n\n```bash\ngit clone https://github.com/photoxpedia/bolor-brain-mcp.git\ncd bolor-brain-mcp\npip install -e .\n```\n\n### 2. Configure Claude Code\n\nAdd to `~/.claude/mcp-config.json`:\n\n```json\n{\n  \"mcpServers\": {\n    \"bolor-brain\": {\n      \"command\": \"python\",\n      \"args\": [\"-m\", \"mcp_server\"],\n      \"cwd\": \"/path/to/bolor-brain-mcp\"\n    }\n  }\n}\n```\n\n### 3. Use\n\n```\n/reason Why is Python popular for data science?\n/debug API returns 500 errors under load\n/decide PostgreSQL or MongoDB for our app?\n/learn-from We fixed the memory leak by increasing connection pool\n```\n\n---\n\n## MCP Tools (11)\n\n### Reasoning (6)\n\n| Tool | What It Does |\n|------|-------------|\n| `reason_hybrid` | Auto-selects best reasoning approach for any query |\n| `reason_symbolic` | Forward/backward chaining with facts and rules |\n| `reason_knowledge_graph` | Graph traversal, path finding, relationship exploration |\n| `reason_case_based` | Find similar past problems and their solutions |\n| `reason_hypothesis` | Generate and test hypotheses from observations |\n| `reason_analogical` | Cross-domain pattern transfer (atom ~ solar system) |\n\n### Memory (4)\n\n| Tool | What It Does |\n|------|-------------|\n| `remember` | Store a case, fact, node, or edge |\n| `recall` | Retrieve matching cases or facts |\n| `learn` | Store problem/solution/outcome (shortcut for remember) |\n| `forget` | Delete a case or fact by ID |\n\n### Utility (1)\n\n| Tool | What It Does |\n|------|-------------|\n| `brain_stats` | Cases, facts, nodes, edges count |\n\n---\n\n## Skills (6)\n\n| Skill | When To Use |\n|-------|------------|\n| `/reason` | Deep analysis of any complex problem |\n| `/debug` | Systematic bug hunting with hypothesis testing |\n| `/decide` | Evidence-based technical decisions |\n| `/learn-from` | Store experiences for future use |\n| `/nsaf` | NSAF evolution integration (requires NSAF MCP) |\n| `/orchestrate` | Meta-orchestration combining Bolor Brain + NSAF |\n\n---\n\n## Persistence\n\nBrain state persists to `~/.bolor-brain/` as JSON:\n\n```\n~/.bolor-brain/\n  cases.json       # Problem -> solution -> outcome\n  facts.json       # Symbolic reasoning facts\n  knowledge.json   # Knowledge graph (nodes + edges)\n```\n\nKnowledge compounds over time. Solve a bug once, recall the solution instantly next time.\n\n---\n\n## With NSAF\n\nAdd NSAF to get evolution capabilities:\n\n```json\n{\n  \"mcpServers\": {\n    \"bolor-brain\": {\n      \"command\": \"python\",\n      \"args\": [\"-m\", \"mcp_server\"],\n      \"cwd\": \"/path/to/bolor-brain-mcp\"\n    },\n    \"nsaf\": {\n      \"command\": \"python3\",\n      \"args\": [\"nsaf_mcp_server.py\"],\n      \"cwd\": \"/path/to/nsaf\",\n      \"env\": { \"PYTHONPATH\": \"/path/to/nsaf\" }\n    }\n  }\n}\n```\n\nTogether: Bolor Brain reasons about WHAT to do. NSAF evolves HOW to do it better. Claude Code executes.\n\nSee [skills/nsaf.md](skills/nsaf.md) and [skills/orchestrate.md](skills/orchestrate.md) for combined workflows.\n\n---\n\n## Testing\n\n```bash\npytest tests/ -v\n# 376 tests\n```\n\n---\n\n## Project Structure\n\n```\nmcp_server.py                    # MCP server (11 tools)\npersistence.py                   # JSON persistence to ~/.bolor-brain/\nmodules/\n  config.py                      # Configuration\n  reasoning_engines/\n    symbolic_reasoner.py          # Forward/backward chaining\n    knowledge_graph.py            # Graph-based knowledge\n    case_based_reasoner.py        # 4R cycle (retrieve, reuse, revise, retain)\n    hypothesis_engine.py          # Hypothesis generation and testing\n    analogical_reasoner.py        # Cross-domain pattern transfer\n    hybrid_reasoner.py            # Orchestrates all 5 engines\nskills/                          # Claude Code skills\n  reason.md, debug.md, decide.md, learn-from.md, nsaf.md, orchestrate.md\ntests/                           # 376 tests\nAGENT_GUARDRAILS.md              # Production safety guidelines\n```\n\n---\n\n## Author\n\n**Bolorerdene Bundgaa**\n- Website: [bolor.me](https://bolor.me)\n- Email: bolor@ariunbolor.org\n\n## License\n\nMIT -- see [LICENSE](LICENSE)\n",
  "bytes": 5117,
  "sha": "206d238dc38db22d07fcdaaf71d2ba9fe4f37dc501c1846d87e03dcb11b53a9d",
  "repo_slug": "photoxpedia/bolor-brain-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_photoxpedia_bolor_brain_mcp_dae9b706/readme"
}