Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Runs a weighted code analysis across eight categories, including error handling, security, naming, and test coverage. Produces a graded report card (A-F) with per-file findings and optional fix suggestions. Works on single files, directories, and git diffs.
Unlike general linters, it specifically targets patterns that emerge when AI-generated code is accepted without review, combining security, naming, and structural checks into a single weighted score.
When to use it
- Auditing a pull request before merging to catch unreviewed AI-generated patterns
- Scoring a freelancer's submitted codebase before accepting delivery
- Checking staged changes for lazy patterns before committing
- Reviewing recent commits on a fast-moving solo project
- Getting a quick quality grade on a directory before a team handoff
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: 'Vibe check my src directory and suggest fixes'
- 1Identifies src/ as the target directory
- 2Runs vibe-check.sh --fix src/ to enable fix suggestions
- 3Parses output for grade, score, per-file breakdown, and category scores
- 4Presents the full Markdown report with unified diff patches
- 5Offers to re-run in --diff mode for focused review of recent changes
Graded report card (A-F) with per-file findings, category weights, and diff patches for each identified issue