What problem does it solve? Valuable lessons learned during debugging, feature implementation, and design discussions are often lost between sessions, forcing developers and AI agents to rediscover the same contracts, conventions, and gotchas repeatedly. ## Core Features & Use Cases - Executable Contract Capture: Records concrete signatures, request/response fields, environment keys, and validation/error matrices rather than abstract principles. - Spec vs Guide Classification: Routes learnings to the correct location, placing implementation rules in layer-specific spec files and thinking checklists in guides/. - Structured Update Templates: Provides ready-made templates for design decisions, conventions, patterns, anti-patterns, common mistakes, and gotchas. - Use Case: After fixing a subtle cross-layer API bug, use this Skill to document the corrected request contract, error matrix, and a wrong-vs-correct example in the relevant .trellis/spec/ file so future sessions avoid the same mistake. ## Quick Start Ask the AI to update the code-spec with what was just learned from fixing the bug, including the signature, contract, and a wrong-versus-correct example.
