Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
A structured workflow agent for development tasks on unfamiliar codebases. It takes a desired outcome, constraints, and available artifacts, then returns ranked findings, an action plan, and concrete checklists. Each step runs through quality gates that require evidence-backed claims and explicit tradeoffs.
Unlike ad-hoc debugging, it enforces evidence-backed claims and explicit fallback plans at each checkpoint, reducing rework from assumptions that turn out wrong.
When to use it
- Onboarding to a new codebase before submitting a pull request
- Planning a risky refactor with fallback steps defined upfront
- Converting messy notes or requirements into a prioritized action plan
- Running pre-release quality checks with defined acceptance criteria
- Delegating a debugging task with a clear owner and ETA for each step
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Map this repo and give me a risk-first plan for adding authentication middleware before Friday
- 1Defines success criteria (middleware integrated, tests passing) and failure criteria (auth bypassed, existing routes broken)
- 2Builds a smallest-valid execution path: identify entry points, locate route definitions, draft minimal middleware stub
- 3Executes in checkpoints, recording file paths, function signatures, and test results as evidence
- 4Applies quality gates: validates assumptions, lists tradeoffs, confirms changes are reversible
- 5Returns a situation summary, ranked findings, a today/this-week action plan, risks with mitigations, and an exact command checklist
A structured report with a 5-line situation summary, impact-ranked findings, a two-horizon action plan, identified risks with mitigations, and a ready-to-run command checklist