janekbaraniewski/openusage/validate-feature
Claude Code and agent skill by janekbaraniewski/openusage
- community
Checking that the SKILL.md file is still in the repository.
Install
npx vskill@latest install janekbaraniewski/openusage/validate-featureSKILL.md
Skill: Validate Feature
Validate that a feature implementation is complete, correct, and ready for review.
When to use
After /implement-feature completes, before /finalize-feature. Also useful standalone to check feature health after manual changes.
Prerequisite
- A design doc in
docs/*_DESIGN.md - Implementation completed (code exists in working tree)
Phases
Phase 0 — Load Context
- Find the design doc for the feature (search
docs/*_DESIGN.mdif path not given). - Read the full design doc. Extract:
- Implementation tasks (Section 7)
- Files listed per task
- Test requirements per task
- Run
git diff main --name-onlyto get list of all changed files. - If an implementation summary exists in conversation context, use it. Otherwise, infer from changed files.
Phase 1 — Build Verification
Run these checks. ALL must pass before proceeding:
make build— binary compiles cleanlymake vet— no vet warnings in changed packagesmake fmt— no formatting issues (checkgofmt -loutput)make lint— no lint errors (skip gracefully if golangci-lint not installed)
Report:
If any check fails, report the error and stop. Do not proceed to Phase 2 with build failures.
Phase 2 — Test Verification
- Identify all Go packages with changed files.
- Run
go test ./<pkg>/... -count=1 -racefor each changed package. - Check for new test files — every task that specifies "Tests:" in the design doc MUST have corresponding test functions.
- Run
go test ./<pkg>/... -count=1 -coverand note coverage for changed packages.
Report:
Flag any design tasks that specify tests but have none implemented.
Phase 3 — Design Compliance
Cross-reference the design doc tasks against actual changes:
- For each implementation task in the design doc:
- Check that ALL files listed under "Files:" were actually modified (or created if new)
- Check that the described functionality exists in the code
- Check that tests specified under "Tests:" exist and pass
- Build a compliance matrix:
- Flag any:
- Tasks with no code changes (skipped?)
- Files changed that aren't in any task (scope creep?)
- Design doc sections marked as "intentional change" or "deferred"
Phase 4 — Code Quality Scan
Scan changed files for common issues:
- Debug artifacts: Search for
fmt.Println,log.Println,FIXME,HACK,XXX,TODO(flag but don't auto-remove TODOs — they may be intentional) - Unused code: Run
go vetwith unused checks. Look for commented-out code blocks. - Error handling: Grep changed files for unchecked errors (bare
errassignments withoutif err != nil). - Import hygiene: Check import grouping follows convention (stdlib, third-party, internal separated by blank lines).
- Secrets/sensitive data: Scan for hardcoded tokens, API keys, passwords. Check no
.envfiles or credentials are staged.
Report:
Phase 5 — Integration Smoke Test
- Run
make buildone final time to confirm clean binary. - If
make demoexists and the feature affects TUI rendering, run it and note if it starts without panics (exit after 2 seconds with timeout). - Check that all changed packages' tests pass together:
go test <all changed packages> -count=1 -race - If the feature added new config fields, verify
configs/example_settings.jsonincludes them.
Phase 6 — Validation Report
Produce a final summary:
If verdict is "NEEDS ITERATION", recommend running /iterate-feature with the issues list.
Rules
- Never auto-fix issues found during validation — report them. Fixing is for
/iterate-feature. - Never skip Phase 1 — build must pass before anything else.
- Never run
go test ./...— only test changed packages unless explicitly asked. - Always cross-reference against the design doc — implementation without design compliance is incomplete.
- If no design doc exists, skip Phase 3 but run all other phases.
- Report findings factually — no opinions on code style unless it violates CLAUDE.md conventions.
Checklist
Before marking validation complete:
- [ ] Build compiles cleanly
- [ ] All changed packages pass tests with -race
- [ ] Every design task has corresponding code changes
- [ ] Every design task with "Tests:" has test functions
- [ ] No debug artifacts in changed files
- [ ] No secrets or credentials in changed files
- [ ] Example config updated if new config fields added
- [ ] Validation report produced with clear verdict
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingdocs/skills/validate-feature/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/janekbaraniewski/openusage/validate-feature)