Agent♥︎Age
Catalog

CPersona

Official

by Cloto-dev · Python

Persistent AI memory in one SQLite file: 3-layer hybrid search, confidence scoring, 29 tools.

CPersona MCP Memory Server (io.github.Cloto-dev/cpersona)

CPersona is a Model Context Protocol (MCP) server that provides persistent AI memory stored in a single SQLite file. It performs 3-layer hybrid search with confidence scoring and exposes a set of tools for memory operations. The implementation is described as having zero LLM dependency.

🛠️ Key Features

  • Persistent memory in one SQLite file
  • 3-layer hybrid search
  • Confidence scoring
  • Tool access: 29 tools (per metadata)

🚀 Use Cases

  • Persisting Claude memory across sessions
  • Memory retrieval using hybrid (semantic/vector) search approaches

⚡ Developer Benefits

  • Works with MCP tooling via the Model Context Protocol
  • Supports SQLite-based storage for reproducible deployments
  • Includes topics aligning with Python, RAG, and semantic/vector search

⚠️ Limitations

  • No additional implementation details are provided beyond the excerpt; full behavior of all tools and the exact search layers are not specified here.

🧩 Topics

ai-agent, claude, llm, mcp, memory, model-context-protocol, persistent-memory, python, rag, semantic-search, sqlite, vector-search

Topics

ai-agentclaudellmmcpmemorymodel-context-protocolpersistent-memorypythonragsemantic-searchsqlitevector-search