From cc-godmode
Orchestrates parallel @validator (TypeScript/unit tests/security) and @tester (Playwright E2E/screenshots/a11y/perf) after @builder, with decision matrix for @scribe or retry.
npx claudepluginhub cubetribe/claudecode_godmode-onThis skill uses the workspace's default tool permissions.
After @builder completes, BOTH quality gates run **simultaneously**.
Enforces quality gates in TypeScript/Node.js projects: pre-commit linting/formatting with ESLint/Prettier/tsc, tests/coverage with Vitest, builds, CI checks, security audits, E2E with Playwright, and Lighthouse performance.
Delivers testing strategies via test pyramid, enforces code quality standards, and validates phase transition gates. Use for testing setup, code reviews, and production readiness.
Runs full validator workflow: executes agent-validate checks via Bash, extracts and reports failures from logs using Task, applies fixes before commit/push/PR. For 'run validator' requests.
Share bugs, ideas, or general feedback.
After @builder completes, BOTH quality gates run simultaneously.
@builder completes
│
┌───────────────┴───────────────┐
▼ ▼
┌─────────────┐ ┌─────────────┐
│ @validator │ │ @tester │
│ Code Quality│ │ UX Quality │
├─────────────┤ ├─────────────┤
│ ✓ TypeScript│ │ ✓ E2E Tests │
│ ✓ Unit Tests│ │ ✓ Screenshots│
│ ✓ Security │ │ ✓ A11y │
│ ✓ Consumers │ │ ✓ Perf │
└──────┬──────┘ └──────┬──────┘
│ │
└───────────────┬───────────────┘
SYNC POINT
│
Apply Decision Matrix
Both agents MUST complete before applying this matrix:
| @validator | @tester | NEXT ACTION |
|---|---|---|
| ✅ APPROVED | ✅ APPROVED | PROCEED to @scribe |
| ✅ APPROVED | 🔴 BLOCKED | RETURN to @builder (with @tester feedback) |
| 🔴 BLOCKED | ✅ APPROVED | RETURN to @builder (with @validator feedback) |
| 🔴 BLOCKED | 🔴 BLOCKED | RETURN to @builder (with MERGED feedback from both) |
Rules:
| Check | Tool | Blocking? |
|---|---|---|
| TypeScript compilation | npx tsc --noEmit | YES |
| Unit tests pass | npm test | YES |
| No security vulnerabilities | Static analysis | YES |
| Consumer impact verified | Type-check consumers | YES if API change |
| Code style / linting | npm run lint | NO (warning only) |
Minimum output: 400 characters Required sections: Summary, Checks Performed, Issues Found, Verdict
| Check | Tool | Blocking? |
|---|---|---|
| E2E tests pass | Playwright MCP | YES |
| Screenshots at 3 viewports | Playwright MCP | YES |
| Console errors captured | Browser console | YES (errors only) |
| WCAG 2.1 AA compliance | a11y checks | YES |
| Core Web Vitals | LCP, CLS, INP, FCP | NO (warning if poor) |
Screenshot viewports:
Minimum output: 800 characters Required sections: Summary, Screenshots Table, Console Errors, Performance Metrics, Verdict
For conflict-free parallel execution, use isolation: worktree:
Task tool → subagent_type: "validator", isolation: "worktree"
Task tool → subagent_type: "tester", isolation: "worktree"
Both agents get their own git worktree — no file conflicts possible.
| Mode | Duration | Improvement |
|---|---|---|
| Sequential | 8–12 min | baseline |
| Parallel | 5–7 min | 40% faster |
| Parallel + Worktree | 5–7 min | 40% faster + zero conflicts |
If an agent crashes (MCP failure, timeout):
Failure report includes: error type, suggested action (retry/escalate/skip), completion percentage.