wreckit

What This Skill Does

wreckit is the definitive code verification engine for the OpenClaw ecosystem, designed to ensure that AI-generated code is not just functional, but bulletproof. Unlike standard AI assistants that often fall into a confirmation bias loop, wreckit enforces a strict, multi-step verification pipeline. It functions as a meta-orchestrator that spawns parallel workers to execute slop-scanning, static type-checking, mutation testing, and regression analysis. By adhering to a rigorous 14-step framework, wreckit ensures that your codebase remains robust, secure, and production-ready, effectively eliminating the common "it looks right but breaks under load" failure mode.

Installation

To integrate wreckit into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/christiancattaneo/wreckit Once installed, the skill automatically hooks into your project's workflow, providing a .wreckit/ directory that serves as your central repository for proof bundles, gate results, and final verification verdicts (Ship/Caution/Blocked).

Use Cases

wreckit is engineered for high-stakes software development. Use it for:

1. Greenfield Projects: Ensure foundational code is bug-free from the first commit.
2. Large-Scale Migrations: Safely port codebases between languages or frameworks with automatic behavior capture and replay.
3. Critical Bug Fixes: Apply patches with total confidence that no regressions are introduced.
4. Security Audits: Scan existing projects against professional-grade SAST and test-quality benchmarks.
5. Verification-First Development: Whenever you require a formal proof bundle to validate that your requirements match the final implementation.

Example Prompts

1. "wreckit: build a Python FastAPI service for user authentication that follows strict REST conventions and includes a full suite of mutation-tested endpoints."
2. "I need to migrate this legacy Java library to TypeScript. Please wreckit the process to ensure 1:1 behavioral parity and provide a final proof bundle."
3. "Fix the memory leak in the data-processing module and run a wreckit verification cycle to prove no regressions were introduced to the pipeline."

Tips & Limitations

• Gate Selection: The skill automatically selects gates based on your intent (BUILD, REBUILD, FIX, or AUDIT). Review the gate configurations if you need custom sensitivity for specific projects.
• Performance: Because wreckit performs deep analysis (including mutation testing), verification cycles can take longer than standard code generation. Plan your compute time accordingly.
• Human Oversight: While wreckit provides a 'Ship' verdict, always perform a final visual audit of the proof bundle logs before deploying to mission-critical production environments.