notque/vexjoy-agent / skills/process/workflow
workflow
VerifiedStructured work: multi-phase tasks, feature builds, planning, objective loops, hill climbing.
references/fl-validate.md
View on GitHub# Feature Validate Phase
Run comprehensive quality gates on the implemented feature. Phase 4 of the feature lifecycle (design > plan > implement > **validate** > release).
## Instructions
### Phase 0: PRIME
1. Read and follow the repository CLAUDE.md before any other action -- it may override defaults for linting, test commands, or gate criteria.
2. Verify feature state is `validate` and `implement` is completed. All state operations go through `python3 ~/.claude/scripts/feature-state.py` -- never modify state files directly.
3. Load implementation artifact from `.feature/state/implement/`. Validation cannot proceed without implementation artifacts; if missing, stop and report.
4. Load L0, L1, and implementation context so quality gates run against the correct scope.
**Gate**: Implementation artifact loaded. Feature state confirms `implement` completed. Proceed.
### Phase 1: EXECUTE (Quality Gates)
**Verification means execution, not reasoning.** Run the command. Do not reason about whether the command would pass. Do not summarize the expected output. Execute the check, paste the exit code, paste the relevant output. A verification phase that produces a verdict without an observed tool result is not a verification — it is a guess with a rigor aesthetic.
**Step 1: Language Detection**
Auto-detect project language(s) from file extensions, build files, and the implementation artifact. This detection drives which quality gate skill runs next.
**Step 2: Run Quality Gates**
Use the repository's existing quality gate skills -- do not re-implement linting or test runners inline. Route to the appropriate skill per language:
| Language | Quality Gate | Command |
|----------|-------------|---------|
| Go | go-patterns | `make check` or `go test ./... && go vet ./...` |
| Python | python-quality-gate | `ruff check && pytest && mypy` |
| TypeScript | universal-quality-gate | `npm run typecheck && npm run lint && npm test` |
| Other | universal-quality-gate | Detect and run project-specific checks |
Show the full, unedited command output for every gate -- never summarize or truncate test results, because summaries hide the exact failure details needed for diagnosis.
**Step 3: Regression Check**
Run the full test suite and compare against pre-implementation baseline:
- New test failures = regression
- Missing tests for new code = coverage gap
Existing tests must still pass; new failures here block advancement regardless of new-feature test results.
**Step 4: Custom Gates**
If the design document specified custom validation criteria, check those too.
**Step 5 (optional): Security and Performance**
If explicitly enabled by the user:
- **Security scan**: Run security-focused review agent
- **Performance check**: Run benchmarks against baseline
These are off by default -- do not run them unless the user requests it.
**Gate**: Every gate must pass. No gate may be skipped and no failure may be approved -- a single failing gate blocks advancement to release. Proceed only when all results are green.
### Phase 2: VALIDATE (Report)
Produce the validation report:
```markdown
# Validation Report: [Feature Name]
## Quality Gates
- [ ] Tests: PASS/FAIL (X/Y passed)
- [ ] Lint: PASS/FAIL (N issues)
- [ ] Type Check: PASS/FAIL
- [ ] Regression: PASS/FAIL
## Coverage
- New files: X
- Modified files: Y
- Test coverage: Z%
## Issues Found
- [Issue 1]: [severity]
- [Issue 2]: [severity]
## Verdict: PASS / NEEDS_FIXES / BLOCK
```
The verdict must reflect actual gate results -- never mark PASS if any gate failed.
If `NEEDS_FIXES`: suggest routing back to the implement phase with specific fix tasks. This phase does not fix failing tests; it reports them and routes back.
If `BLOCK`: explain blocking issues.
**Gate**: Report produced. Proceed to Checkpoint.
### Phase 3: CHECKPOINT
1. Save validation artifact (all state operations through the feature-state script):
```bash
echo "VALIDATION_REPORT" | python3 ~/.claude/scripts/feature-state.py checkpoint FEATURE validate
```
2. If validation passed, advance:
```bash
python3 ~/.claude/scripts/feature-state.py advance FEATURE
```
3. Suggest next step:
```
Validation passed. Run /feature-lifecycle to continue to the release phase.
```
## Error Handling
| Error | Cause | Solution |
|-------|-------|----------|
| Tests fail | Implementation bugs | Route back to implement phase with failure details |
| No test framework detected | Project setup incomplete | Report gap, suggest setup |
| Lint failures | Style issues | Auto-fix if trivial, otherwise route back |
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| planning-and-task-breakdownaddyosmani/agent-skillsBreaks work into ordered tasks. Use when you have a spec or clear requirements and need to break work into implementable tasks. Use when a task feels too large to start, when you need to estimate scope, or when parallel work is possible. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| screenpipethedotmack/claude-memSearch screen and audio history recorded by Screenpipe and optionally save selected context into claude-mem. Use when the user asks to recall desktop activity, a meeting, or something seen in another app through Screenpipe. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| contact-widgetnexu-io/open-designSelf-contained floating chat widget with welcome screen, social links, meeting button, and message input. Single HTML file, zero dependencies. | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |
| meeting-notesnexu-io/open-designMeeting notes page — title bar with attendees, agenda checklist, decisions block, action items table with owners + dates, and a "next meeting" footer. Use when the brief mentions "meeting notes", "minutes", "1:1 notes", "all-hands recap", or "会议纪要". | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |