ai-self-evolution

What problem does it solve?

It solves the problem of losing hard-won experience from failed commands, user corrections, and outdated knowledge so the system can improve across sessions instead of repeating the same mistakes.

Core Features & Use Cases

• Captures learning in a structured loop: records errors, corrections, feature requests, and best practices into the project .learnings/ markdown logs for later review.
• Session-level recap and closure: performs an automatic “session start” pending/high-priority scan and a “session end” summary to mark resolutions and add missing entries.
• Detects triggers and strengthens recurring patterns: logs when non-obvious issues are found, when tools/APIs fail, and when the same pattern repeats, using Pattern-Key and recurrence counts to guide promotion or skill extraction.

Quick Start

Ask the AI to complete a task, and if it fails or you correct it, instruct it to log the resulting learning into the project .learnings/ files following this skill’s formats.