agent-memory-skills

What problem does it solve?

Agents operate with static configurations and limited introspection. This Skill enables continuous self-improvement by storing experiences, evaluations, and performance metrics in ChromaDB, while isolating memory from static .md files.

Core Features & Use Cases

• Persistent learning after each task: remember patterns, preferences, and improvements for future tasks.
• Automatic self-evaluation: measure quality, time, and outcomes to drive better behavior.
• Structured memory: separate collections for improvements, evaluations, and performance; supports global and project-scoped memory.
• Use Case: a development agent that improves its code-review process over time by recalling past insights and tracking performance.

Quick Start

Initialize the agent-memory framework and trigger a sample task to start learning and evaluating automatically.