What problem does it solve?
This skill defines AI validation criteria for work in progress, enabling explicit verification standards when a spec or plan lacks clear validation.
Core Features & Use Cases
- Default validation checklist: Introduces the standard validation items (mocked time, mocked network, mocked filesystem, mocked LLM calls, 90%+ code coverage, and code review) to ensure consistent quality.
- Custom validation prompts: Allows adding feature-specific verification requirements beyond the defaults.
- Standalone and appendable outputs: Provides options to append to an existing spec/plan or save as a standalone lore document in .lore/validation/[feature-or-work].md.
- Context-rich guidance: Keeps a trace of how the criteria were derived for future audits.
Quick Start
Use this skill to generate AI validation criteria for your current work item. If a spec/plan exists, offer to append the AI Validation section; if not, save as a standalone document in .lore/validation/[feature-or-work].md.