Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Forge is an autonomous quality engineering swarm that runs E2E tests against a real backend, analyzes failures, applies tiered fixes, and commits only when all 7 quality gates pass. It combines BDD/Gherkin behavioral specs with DDD/ADR/TDD methodology across any architecture. No mocking or stubbing is allowed — all tests run against a live backend.
It closes the loop between test failure and verified commit automatically, running spec generation, root cause analysis, tiered fixing, and 7 gate enforcement without manual intervention.
When to use it
- Auto-fixing a failing payments context test suite before a release
- Generating Gherkin specs for an undocumented microservice bounded context
- Detecting API contract regressions after a backend schema change
- Running chaos tests to verify offline and timeout handling in a mobile app
- Tracking fix pattern confidence across repeated CI failures in a monorepo
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Run Forge on the payments context
- 1Check backend health endpoint and start the backend if not running
- 2Verify Gherkin specs exist for payments or generate them from implementation files
- 3Run E2E tests against the live backend API, prioritizing predicted-to-fail tests
- 4Analyze failures by root cause category and apply confidence-tiered fixes
- 5Evaluate all 7 quality gates (functional, behavioral, coverage, security, accessibility, resilience, contract)
All payments tests passing with a detailed commit message listing gate statuses, root causes, and fix pattern confidence tiers