Thezenmonster/agentmem

What it does

Thezenmonster/agentmem is a memory management system designed specifically for coding agents. It provides a sophisticated governed memory framework that tracks information through a trust lifecycle: hypothesis → active → validated → deprecated. The system includes built-in conflict detection to identify contradictory information, staleness tracking to monitor data freshness, and health scoring to assess overall memory quality.

How to set up

Installation is straightforward via pip: pip install quilmem[mcp]. The tool operates with zero infrastructure requirements, leveraging SQLite with FTS5 (Full-Text Search 5) for efficient storage and retrieval. Simply install the package and integrate it with your MCP-compatible coding agent environment. It works across Python, macOS, Windows, and Linux platforms.

Key features

• Trust lifecycle management (hypothesis → active → validated → deprecated states)
• Automatic conflict detection between stored memories
• Staleness tracking to monitor information age and relevance
• Health scoring system for memory quality assessment
• SQLite + FTS5 backend for efficient full-text search
• Zero infrastructure deployment - works locally
• MCP server compatible for integration with AI agents