MindCore Memory MCP

AI Long-Term Memory Server — Production-grade persistent memory for AI agents.

"The best AI agent isn't the smartest — it's the one that remembers."

Value Proposition

MindCore Memory solves AI Agent's biggest pain point: limited context windows, lost information in long conversations, and broken cross-session memory continuity.

What Problem It Solves

Pain Point	Status Quo	MindCore Memory
AI forgets everything	Conversation ends, all lost	Persistent long-term memory
No cross-session recall	Re-teach every session	Cross-session knowledge reuse
Memory chaos, no priority	All memories weighted equally	Importance grading + confidence
RAG brute-force injection	Context overload, quality drops	Precise context window

Quick Start (3 lines)

# 1. Install
pip install mindcore-memory

# 2. Launch MCP Server
mindcore-memory

# 3. Call from your AI Agent
memory_id = memory_store("User says his name is Zhang San, free on Wednesday")
context = memory_recall("User's schedule")

Eval Framework Results

Storage Integrity:     100% (data persistence correct)
Recall Relevance:      100% (relevant memories recalled first)
Confidence Calibration: 100% (confidence correctly calibrated)
Importance Weighting:   100% (high-priority memories ranked higher)
Context Efficiency:    100% (context window not overloaded)

Overall Score: 100%

Core Tools

memory_store - Store memory

memory_store(
    content="Python was created by Guido van Rossum from Netherlands",
    importance=3,        # 1-4 importance level
    tags=["python", "history"],
    confidence=0.95,      # confidence score
    source="agent"       # agent/user/tool
)

memory_recall - Recall memory

memory_recall(
    query="Who created Python",
    tags=["python"],      # optional tag filter
    limit=10             # return count
)

memory_context - Build context window

# Build optimal context for current task (auto-dedup + priority sort)
context = memory_context(
    query="Current project status",
    max_tokens=2000      # auto-truncate
)

memory_stats - System status

# View memory statistics: total/distribution/confidence
stats = memory_stats()

Project Structure

mindcore-memory-mcp/
├── mindcore_memory/          # Python package (pip install entry)
│   ├── __init__.py
│   ├── memory_engine.py      # Core memory engine
│   ├── server.py             # MCP Server (stdio + HTTP dual transport)
│   ├── http_app.py           # HTTP endpoint (production deploy)
│   └── eval_framework.py     # Evaluation framework
├── tests/
│   └── test_memory.py        # Unit tests
├── examples/
│   └── basic_usage.py        # Usage examples
├── pyproject.toml
├── README.md
└── LICENSE

Integration

Claude Desktop

{
  "mcpServers": {
    "mindcore-memory": {
      "command": "pip",
      "args": ["install", "mindcore-memory"]
    }
  }
}

VS Code AI

Search MindCore Memory in the extension marketplace.

HTTP API (Production)

curl -X POST http://localhost:8080/mcp \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -d '{"jsonrpc":"2.0","method":"tools/call","params":{"name":"memory_store","arguments":{"content":"test"}},"id":1}'

Production Standards

Standard	Implementation
JSON-RPC 2.0	stdio + HTTP dual transport
Bearer Token Auth	Optional auth for HTTP endpoints
Input Validation	Pydantic schemas
CI/CD	GitHub Actions
Unit Tests	pytest + coverage
Eval Framework	5 core metrics
Observability	structlog complete logging
Data Sovereignty	JSONL local files, no vendor lock-in

Open Source

This project is open source (MIT License). The code is completely free. Storage uses local JSON files with no cloud service dependency and no data collection.

License

MIT License - see LICENSE file for details.

---

Star History

---

<p align="center"> <strong>Give AI memory. Make humans trust AI more.</strong> </p>

---

Collaborate With Me

I'm actively developing AI safety architecture projects including:

1. Cerebellum Evolution Engine - AI safety & evolution framework
2. MindCore - Cognitive memory architecture for AI systems
3. Border Guard - Self-evolving security operating system
4. Ternary Balance Boundary Algorithm - Novel equilibrium theory with 3 papers
5. MindCore Memory MCP - This project: production-grade long-term memory server

Interested in collaborating? Reach out:

• Email: 1410770089@qq.com
• GitHub: @woshilaohei

Author: Lao Hei