How to handle /recap vs /status overlap? Options: merge, keep separate with clearer scopes, or remove one · Phase 1
User Tasks
Test each skill after fixes are applied · Phase 1
Summary
Review all existing skills for correctness, usefulness, and consistency. Fix issues found.
Problem / Motivation
Skills were created early and have accumulated issues: wrong field values, references to non-existent dashboard sections, overlap between skills, and scaling problems.
Proposed Solution
Audit each skill, document issues, fix them in one pass.
Open Questions
No open questions.
Phase Overview
Phase
Description
Status
Phase 1
Fix known issues per skill
—
Phase 2
Manual test all skills
—
Phase 1: Fix Known Issues —
Goal: Fix all concrete issues found during review.
Inventory (7 skills)
Skill
Status
Issues
/status
needs fix
Asks for “Next Actions” section that doesn’t exist in feature-dashboard. Capabilities step adds little value (always “pre-code phase”).
/recap
ok
Works well. Overlaps with /status (both show commits, inbox, proposals) — needs scoping decision.
/new-feature
needs fix
Uses status: inbox (doesn’t exist, should be new). References tags field not in template. Dashboard update step is fragile at 64+ FRs.
/idea
ok
Clean and compact. Clear distinction from /new-feature.
/inbox-review
ok
Solid flow, handles empty inbox.
/feature-review
needs fix
No limit or filter — will try to review all 64 FRs in 02_new/ one by one. Needs batch size or priority filter.
/accept-proposal
ok
Works correctly.
Fixes
Fix
Skill
Details
Owner
Status
Remove “Next Actions” reference
/status
Step 2 asks for section that doesn’t exist
opus
—
Remove or simplify capabilities step
/status
Step 1 output is always the same, low value
opus
—
Fix status: inbox → new
/new-feature
Align with actual lifecycle statuses
opus
—
Remove tags field
/new-feature
Field doesn’t exist in feature-request template
opus
—
Simplify dashboard update
/new-feature
Don’t update 2 dashboards inline — too fragile
opus
—
Add batch limit
/feature-review
Show max 5 FRs per run, sorted by priority/FR number
opus
—
Phase 2: Manual Test —
Goal: Run each skill and confirm it works end-to-end.
Skill
Test
Owner
Status
/status
Run and verify output matches reality
mv
—
/recap
Run at session start, verify all sections
mv
—
/new-feature
Create a test FR, verify file + frontmatter
mv
—
/idea
Capture a test idea, verify file
mv
—
/inbox-review
Add a note to quick-notes.md, run skill
mv
—
/feature-review
Run with FRs in 02_new/, verify limit works
mv
—
/accept-proposal
Create a test proposal, accept it
mv
—
Prerequisites / Gap Analysis
Current State
Component
Status
Details
/new-feature
broken
Wrong status value, phantom tags field
/status
broken
References non-existent dashboard section
/feature-review
poor
Doesn’t scale with 64 FRs
/idea
ok
Works as expected
/inbox-review
ok
Works as expected
/accept-proposal
ok
Works as expected
/recap
ok
Works, needs scoping vs /status
Test
Manual tests
Test
Expected
Actual
Last
Run each skill after Phase 1 fixes
Completes without errors, correct output
pending
-
/new-feature creates FR with status: new
Frontmatter has status: new, no tags field
pending
-
/feature-review shows max 5 FRs
Only 5 shown, sorted by priority
pending
-
/status output matches actual system state
All counts and sections correct
pending
-
AI-verified tests
Scenario
Expected behavior
Verification method
/new-feature output matches template
Generated file has all required frontmatter fields, no extras
Parse frontmatter, compare to template
/status references only existing dashboard sections