Skip to content

PR Review

Review a software change (pull request) against accumulated acceptance criteria, tests-as-evidence, and project conventions. Use when evaluating a PR from an agent or human contributor - decide what to bounce, what to fix inline, and what to approve.

Software Engineering|v1|Updated 7/16/2026|License: Proprietary
MCP get_skill({ skillId: "pr-review" })

Use this skill with your agent

Create a free account and connect via MCP

Get Started Free
---
title: "PR Review"
description: "Review a software change (pull request) against accumulated acceptance criteria, tests-as-evidence, and project conventions. Use when evaluating a PR from an agent or human contributor - decide what to bounce, what to fix inline, and what to approve."
author: AgentArmory
license: Proprietary
---

# PR Review

Review software changes against criteria, not intuition. Every PR review follows the same structure: verify the tests prove the behavior, confirm every acceptance criterion is met, check for regressions, enforce project conventions, and decide the disposition. No guessing, no taste-based preferences.

Research shows AI-generated PRs have ~1.7x more issues than human-written PRs (CodeRabbit, 2025), and that traditional "lightweight" PR review is insufficient for AI-generated code (Metacto, 2026: 10-point review checklist). This skill provides a structured review protocol calibrated for the agent era.

## Table of Contents

- [When to Use](#when-to-use)
- [Triggers](#triggers)
- [Prerequisites](#prerequisites)
- [Methodology](#methodology)
- [Dos](#dos)
- [Don'ts](#don'ts)
- [Pitfalls](#pitfalls)
- [Nonsense Check](#nonsense-check)
- [Validate Before Shipping](#validate-before-shipping)
- [Cross-Skill Hints](#cross-skill-hints)

## When to Use

Use when a pull request, diff, or change set is submitted for review. This includes both PRs from other agents and your own PR before submitting (self-review is the fastest review cycle).

Do NOT use for: discussion-only reviews (use the project's RFC process), design reviews (use the project's architecture review), or configuration-only changes (verify the config file is valid).

## Triggers

"review this PR", "code review", "review this change", "review my PR", "please review", "check this diff", "review", "approval needed", "PR ready for review", "review this diff", "review before merge"

## Prerequisites

- The diff or PR URL accessible (local git diff, GitHub PR URL, or inline patch)
- The project's testing conventions, style guide, and commit conventions
- The original acceptance criteria or spec that motivated the change
- The full test suite pass results for the target branch
- CI check results (if available)

## Methodology

### Phase 1: Read the Spec First

Before looking at a single line of code, re-read the acceptance criteria or spec that motivated this change. This is the most important step reviewers get wrong - they review against taste rather than against requirements.

Continue reading

Sign up for a free account to view the full skill content

Login / Register
#code-review#pr-review#quality-gate#acceptance-criteria#tests-as-evidence#software-engineering
License: Proprietary

Related Skills

More skills in Software Engineering

Accessibility Standards

Comprehensive web accessibility standards based on WCAG 2.2 AA, with 38+ anti-patterns, legal enforcement context (EAA, ADA Title II), WAI-ARIA patterns, and framework-specific fixes for modern web frameworks and libraries.

#github-copilot#accessibilityMIT

Accord

Authoring unified specification packages across Business/Development/Design teams via staged elaboration (L0 Vision → L1 Requirements → L2 Team Detail → L3 Acceptance Criteria). No code. Use when authoring cross-team specs, building L0-L3 packages, or aligning Biz/Dev/Design on a single source of truth.

#broad-capability#developmentMIT

Acquire Codebase Knowledge

Use this skill when the user explicitly asks to map, document, or onboard into an existing codebase. Trigger for prompts like "map this codebase", "document this architecture", "onboard me to this repo", or "create codebase docs". Do not trigger for routine feature implementation, bug fixes, or narrow code edits unless the user asks for repository-level discovery.

#github-copilot#documentationMIT

Acreadiness Assess

Run the AgentRC readiness assessment on the current repository and produce a static HTML dashboard at reports/index.html. Wraps `npx github:microsoft/agentrc readiness` and hands off rendering to the @ai-readiness-reporter custom agent. Supports policies (--policy) for org-specific scoring. Use when asked to assess, audit, or score the AI readiness of a repo.

#github-copilot#planningMIT

Acreadiness Generate Instructions

Generate tailored AI agent instruction files via AgentRC instructions command. Produces .github/copilot-instructions.md (default, recommended for Copilot in VS Code) plus optional per-area .instructions.md files with applyTo globs for monorepos. Use after running /acreadiness-assess to close gaps in the AI Tooling pillar.

#github-copilot#skillMIT

Acreadiness Policy

Help the user pick, write, or apply an AgentRC policy. Policies customise readiness scoring by disabling irrelevant checks, overriding impact/level, setting pass-rate thresholds, or chaining org baselines with team overrides. Use when the user asks about strict mode, AI-only scoring, custom weights, CI gating, or wants org-wide standardisation.

#github-copilot#planningMIT
PR Review - AgentArmory Skill — AgentArmory