myna-dev-coveragelisted

# Myna Dev Coverage Audit Myna skills for feature coverage — does every feature from `docs/features/*.md` have executable instructions in its owning skill? Not "mentioned" — actually covered with steps Claude would follow. ## Arguments $ARGUMENTS Parse for: - No arguments: all feature skills + steering + main + config examples - Specific paths/globs: check only matching skills (always read ALL feature spec files regardless) - `--uncommitted`: only skills with uncommitted changes - Anything else: error out ## Setup Read ALL of these before auditing: 1. `docs/design/architecture.md` — extract "Features covered:" for every skill. Build complete map: skill → features. 2. All feature spec files in `docs/features/*.md` 3. `docs/design/foundations.md` — vault structure, canonical templates 4. Every skill file in scope ## Coverage Grades - **FULL** — executable read → decide → write steps for every meaningful behavior. Minor defaults Claude handles are fine. - **PARTIAL** — core behavior covered but specific behaviors/edge cases the spec explicitly calls out are missing. - **NONE** — assigned in architecture but no executable steps in the skill body. ## What "Executable" Means A sub-feature has executable steps if the skill has: - Specific file paths to read - Specific data to extract or decisions to make - Specific files/sections to write to - Concrete output format NOT executable: named in a comment, described only in a worked example, covered by a vague verb. ## Method