code-review

Assesses GitHub pull requests for defects and produces scored, actionable verdicts.

Updated Apr 27, 2026
One-click install
npx skills add https://github.com/vTRKA/supervibe --skill code-review-vtrka
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-review
Source: https://github.com/vTRKA/supervibe/tree/main/skills/code-review
Command: npx skills add https://github.com/vTRKA/supervibe --skill code-review-vtrka

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It prevents risky or incorrect changes from being merged by forcing a structured, evidence-backed review across correctness, security, and other critical quality dimensions.

Core Features & Use Cases

  • Evidence-first pre-merge review: Ranks findings (CRITICAL/MAJOR/MINOR/SUGGESTION) and blocks when required source, approval, validator output, or runtime receipt proof is missing.
  • Special handling for refactors and protected simplification blocks: Enforces symbol/caller evidence checks and blocks approvals when simplification markers are malformed or overlap protected ranges.
  • Actionable output for the code-review handoff: Produces a consistent decision report with evidence, confidence scoring requirements, and a clear nextAction.

Quick Start

Before you merge, ask your AI tool to run the code-review workflow for the PR and return a severity-ranked verdict with exact evidence and the nextAction if anything is missing.

Frequently Asked Questions about code-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I systematically review code changes before a pull request merge?

To review code changes before a merge, perform an evidence-backed assessment across correctness, security, and compatibility dimensions. Execute typecheck, tests, and lint, then rank findings by severity to determine merge readiness and prevent regressions.

What is evidence-backed code review and how does it prevent merge regressions?

Evidence-backed code review prevents risky merges by requiring validator output, runtime receipts, and approval proof before allowing changes. It systematically blocks approvals when required source evidence is missing, ensuring proposed changes do not introduce regressions.

How do I validate refactor safety and protected blocks during a pull request review?

To validate refactor safety during a pull request review, enforce symbol and caller evidence checks. Block approvals when simplification markers are malformed or overlap protected ranges, ensuring the refactoring does not break existing public symbols.

Can I use static validation to rank code review findings by severity?

Yes, you can use static validation to rank code review findings by severity. Categorize issues as CRITICAL, MAJOR, MINOR, or SUGGESTION based on typecheck and lint output, then enforce stop conditions when evidence is missing.

What is the best way to enforce stop conditions when approval evidence is missing?

The best way to enforce stop conditions for missing approval evidence is to block the merge process automatically. Halt the review workflow and emit a structured decision report with confidence scoring and a clear nextAction.

Does code review handle compatibility checks for public symbols and working-path boundaries?

Yes, code review handles compatibility checks by validating changes across public symbols, protected blocks, and working-path boundaries. It reads change scope and source-of-truth conventions to prevent compatibility regressions before merging.