ultraqa

UltraQA is an advanced autonomous testing and quality assurance skill designed for Claude Code. It orchestrates a multi-agent feedback loop to ensure software reliability without requiring constant human intervention. By wrapping development tasks in a rigorous test-verify-fix cycle, it automates the resolution of regressions, build failures, linting issues, and TypeScript errors.

• Autonomous QA Cycle: Orchestrates a sequence of qa-tester, architect, and executor agents to handle failures iteratively.
• Goal-Oriented Execution: Supports specific test modes (--tests, --build, --lint, --typecheck) or user-defined custom patterns.
• Intelligent Diagnostics: Leverages architect-level reasoning to analyze test failure outputs and generate precise fix recommendations.
• State Persistence: Tracks progress across cycles in .omc/ultraqa-state.json to ensure recovery and reliability if sessions are interrupted.
• Advanced Exit Logic: Implements safety guardrails by exiting early if the same failure persists or if the maximum cycle limit (default 5) is reached.

Ideal for developers seeking to automate complex verification pipelines, UltraQA takes the burden out of TDD and CI/CD maintenance. It is particularly effective for large codebases where multi-file changes require systematic verification. Users simply provide a goal, and the skill handles the heavy lifting of running commands, diagnosing stack traces, and applying fixes. Note that this skill requires a clean environment and respects workspace state management, ensuring that state files are scrubbed upon completion or cancellation. It serves as a powerful extension to the core OMC orchestration suite for maintaining high-quality software standards.