learn-from-mistakes
CommunityLog retrospectives to improve future work.
System Documentation
What problem does it solve?
Runs after human sign-off on a chunk. Logs human corrections, missed cases, and implementation holes for pattern tracking. Does not fix anything — the goal is to accumulate insight over time. When the same class of mistake occurs 3 or more times, surfaces a concrete suggestion to update AGENTS.md or an existing skill. Trigger after a human marks a chunk complete, when the user says "log what went wrong", "retrospective", "what did we miss", or "learn from this".
Core Features & Use Cases
- Logs events after sign-off to capture corrections, missed cases, and implementation gaps for pattern tracking.
- Aggregates patterns to surface concrete suggestions for updating AGENTS.md or improving skills.
- Writes learning entries to .claude/learnings.md to enable trend analysis and future improvements.
Quick Start
Log retrospective entries after sign-off to capture corrections and patterns for future improvement.
Dependency Matrix
Required Modules
None requiredComponents
Standard package💻 Claude Code Installation
Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.
Please help me install this Skill: Name: learn-from-mistakes Download link: https://github.com/rolopez23/agent-harness/archive/main.zip#learn-from-mistakes Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
Agent Skills Search Helper
Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.