What problem does it solve?
When developers fix non-obvious bugs, make architectural decisions, or update interface contracts, the context and insights behind these resolutions often fade once the work is done, leading to repeated mistakes, wasted investigation time, and inconsistent project knowledge across team members and AI coding agents.
Core Features & Use Cases
- Judgment-Gated Filtering: Automatically filters out trivial, non-reusable changes to only record high-value lessons that would change a future agent's approach to a similar problem.
- Commit-Anchored Verification: Enforces that every lesson is tied to a verified 40-character git commit hash, ensuring all recorded insights are tied to proven, working solutions.
- Standardized Lesson Formatting: Provides a mandatory template with sections for situation, root cause, actionable rules, and application boundaries, so lessons are readable and actionable for future agents reading them without context.
- Use Case: After your team spends 3 hours debugging a non-obvious race condition in the payment processing flow, use this skill to record the root cause, avoidance rule, and commit hash so no future agent wastes time on the same issue.
Quick Start
Use the distilling-lessons skill to record the root cause and avoidance rule for the recent authentication timeout bug you fixed, anchored to its verified git commit.
