tpo-code-review-checklist

Reviews code changes against requirement, architecture, security, and testing checklists with structured severity decisions.

Updated Mar 22, 2026
One-click install
npx skills add https://github.com/TECH-HY/SKILLS --skill tpo-code-review-checklist-tech-hy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tpo-code-review-checklist
Source: https://github.com/TECH-HY/SKILLS/tree/main/skills/tpo-code-review-checklist
Command: npx skills add https://github.com/TECH-HY/SKILLS --skill tpo-code-review-checklist-tech-hy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Code reviews often become inconsistent, subjective, or superficial, letting security flaws, untested logic, and requirement mismatches slip into production. This Skill gives a Technical Product Owner agent a rigorous, repeatable review framework with clear decision types and severity levels. ## Core Features & Use Cases - Structured Review Decisions: Classify every review as Approve, Request changes, Comment, or Nit, with Blocker/Must-fix/Should-fix/Nit severity levels. - Eleven-Dimension Checklist: Covers requirement fit, architecture, readability, correctness, tests, security, database safety, API quality, UX, performance, and documentation. - Standardized Output Template: Produces a consistent Code Review Result report with blockers, must-fix issues, and a concrete correction task for the developer agent. - Use Case: An AI Developer Agent submits a pull request for a new API endpoint. The TPO agent runs this checklist, flags missing authorization checks as a Blocker, notes weak test coverage as Must-fix, and returns a structured correction task instead of a vague rejection. ## Quick Start Review the attached pull request using the code review checklist and return a structured decision with blockers, must-fix issues, and a correction task.

Frequently Asked Questions about tpo-code-review-checklist

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

FAQPage Schema
How do I structure a code review checklist for AI-generated code?

Use a checklist covering requirement fit, architecture, correctness, tests, security, and UX, with explicit severity levels like Blocker and Must-fix. This Skill provides exactly that structure plus a standardized output template for consistent review decisions.

What severity levels should a code review use?

This checklist defines four levels: Blocker (cannot merge, e.g., security vulnerability or failing tests), Must fix (correct before acceptance), Should fix (recommended improvement), and Nit (optional polish that never blocks delivery).

What inputs are required before reviewing a pull request?

Collect the original task, acceptance criteria, affected files, implementation and test summaries, known limitations, screenshots for UI changes, and migration or API contract notes. If these inputs are missing, request them before approving the work.

When should a code review reject instead of approve?

Reject when requirements are unmet, security is uncertain in sensitive flows, business logic is untested, authorization is unclear, secrets are hardcoded, or database changes lack migration review. Every rejection must convert into a clear correction task.

Does this checklist cover database and API changes?

Yes. It includes dedicated sections for data and database safety (migrations, indexes, transactions, rollback) and API quality (schemas, error consistency, status codes, pagination, idempotency, and versioning of breaking changes).