code-review

Review entire Gait repository for regressions, safety, and contract correctness.

12|5|Updated Feb 5, 2026
One-click install
npx skills add https://github.com/davidahmann/gait --skill code-review-davidahmann
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-review
Source: https://github.com/davidahmann/gait/tree/main/.agents/skills/code-review
Command: npx skills add https://github.com/davidahmann/gait --skill code-review-davidahmann

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill addresses the need for a comprehensive, risk-focused code review that goes beyond typical pull request diffs, ensuring the entire codebase is audited for regressions, safety vulnerabilities, and contract adherence.

Core Features & Use Cases

  • Full-Repository Analysis: Reviews the entire codebase, not just recent changes.
  • Risk Prioritization: Identifies and ranks findings based on severity (P0-P3), focusing on safety, determinism, and contract correctness.
  • Use Case: Before releasing a new version of the Gait tool, use this Skill to perform a thorough audit to catch potential regressions or safety issues that might have been introduced across the entire project.

Quick Start

Use the code-review skill to review the entire repository and report severity-ranked technical risks.

Frequently Asked Questions about code-review

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

FAQPage Schema
How do I audit my entire codebase for regressions and safety risks?

To audit your codebase for regressions, perform a full-repository code review that analyzes high-risk surfaces like core modules, CLI behavior, and schemas, reporting findings ranked by severity.

What is a full-repository code review and when do I need it?

A full-repository code review evaluates your entire codebase for safety vulnerabilities, determinism breaks, and contract drift, needed before major releases to catch regressions missed by typical pull request diffs.

How do I check my CLI and SDK boundaries for contract correctness?

Checking CLI and SDK boundaries for contract correctness involves analyzing adapter boundaries and documentation for contract drift, safety bypasses, and false-green paths to report severity-ranked technical risks.

Does full-repository code review work for pre-release risk assessment?

Full-repository code review works for pre-release risk assessment by scanning core modules, schemas, and documentation to identify safety bypasses, portability assumptions, and reproducibility breaks before deployment.

What's the best way to identify fail-closed safety bypasses across a project?

The best way to identify fail-closed safety bypasses is to execute a Codex-style repository audit that reviews core modules and CLI behavior for integrity weakening and false-green paths, ranking findings by severity.

How do I find documentation inaccuracies and portability assumptions in my repo?

Finding documentation inaccuracies and portability assumptions requires auditing the entire repository, analyzing schemas and documentation surfaces for contract drift and reproducibility breaks to report technical risks.