Run any Skill in Manus with one click

Get Started

pr-review

Stars0

Forks0

UpdatedJuly 15, 2026 at 21:59

Cognitive pre-merge linter — analyzes full diff against conventions and best practices

Installation

Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.

Run Skill in Manus

Source

gonzoblasco

gonzoblasco/workflow-kit

View GitHub Repository View Creator Repositories

Download

Run Skill in Manus

Related occupationsSOC

Based on SOC occupation classification

Software Quality Assurance Analysts and TestersComputer and Mathematical Occupations·SOC 15-1253

SKILL.md

readonly

name	pr-review
description	Cognitive pre-merge linter — analyzes full diff against conventions and best practices
version	2.0.0
compat	claude-code, codex, cursor, openclaw

Skill: PR Review (Cognitive Pre-Merge Linter)

Purpose

Act as an automated second pair of eyes. Analyzes the full branch diff (git diff main...HEAD) against repository conventions and engineering best practices before consolidating code.

Audit Dimensions

1. Functional Correctness and Edge Cases

Look for unhandled error paths (promises without .catch or try/catch, null/undefined values that could break the flow).
Detect potential race conditions or inconsistent client/server states.

2. Test Coverage

Validate that core business logic or new endpoints have corresponding test files.
Lack of coverage in critical components is considered blocking.

3. Conventions, Architecture, and Cleanliness

Contrast changes against the repository's architectural definitions.
Detect residual console.log, debug comments (// TODO: remove this), or dead code.

4. Early Layer Security

Secrets: Strictly prohibit hardcoded credentials, tokens, or API keys.
Injections: Verify that user inputs are typed, validated, or sanitized before touching databases or system operations.
Auth guards: Ensure every new route has its corresponding middleware or session validation.

5. Efficiency and Performance

Identify unnecessary iterations, heavy main-thread mutations, or inefficient database queries (e.g., avoidable N+1 pattern).

Strict Output Report Structure

The agent must render the review using exactly this Markdown template:

## 🔍 Code Review Report — [Branch/Task]

### ✅ Positives

- [Design strengths, modularity, or good practices applied]

### 🔴 Blocking Findings (Require Mandatory Fix)

- **[File:Line]** — [Description of security, build, or severe logic issue]

### 🟡 Suggested Improvements (Do Not Block Merge)

- **[File:Line]** — [Readability recommendation, minor optimization, or future refactor]

### 🔵 Technical Notes / Debt

- [Contextual notes on architecture or long-term behavior]

### 🛠️ FINAL VERDICT

[APPROVED / REQUEST CHANGES]

More from this repository

same repository

product-reset

gonzoblasco/workflow-kit

Reset estratégico de un producto existente: diagnosticar, visionar, planificar, ejecutar

2026-07-160

close-epic

gonzoblasco/workflow-kit

Formal epic closure ritual — security audit, CHANGELOG consolidation, tagging, merge

2026-07-150

commit-formatter

gonzoblasco/workflow-kit

Format commit messages following Conventional Commits — works in terminal and IDE

2026-07-150

commit

gonzoblasco/workflow-kit

Format and create commits following Conventional Commits with project-specific scopes

2026-07-150

copilot

gonzoblasco/workflow-kit

Universal workflow orchestrator — Pre-Flight checklist + 5 Gates for task lifecycle

2026-07-150

doc-writer

gonzoblasco/workflow-kit

Generate and update technical documentation — architecture, conventions, design decisions, changelogs

2026-07-150

Skill: PR Review (Cognitive Pre-Merge Linter)

Purpose

Act as an automated second pair of eyes. Analyzes the full branch diff (git diff main...HEAD) against repository conventions and engineering best practices before consolidating code.

Audit Dimensions

1. Functional Correctness and Edge Cases

Look for unhandled error paths (promises without .catch or try/catch, null/undefined values that could break the flow).

Detect potential race conditions or inconsistent client/server states.

2. Test Coverage

Validate that core business logic or new endpoints have corresponding test files.

Lack of coverage in critical components is considered blocking.

3. Conventions, Architecture, and Cleanliness

Contrast changes against the repository's architectural definitions.

Detect residual console.log, debug comments (// TODO: remove this), or dead code.

4. Early Layer Security

Secrets: Strictly prohibit hardcoded credentials, tokens, or API keys.

Injections: Verify that user inputs are typed, validated, or sanitized before touching databases or system operations.

Auth guards: Ensure every new route has its corresponding middleware or session validation.

5. Efficiency and Performance

Identify unnecessary iterations, heavy main-thread mutations, or inefficient database queries (e.g., avoidable N+1 pattern).

Strict Output Report Structure

The agent must render the review using exactly this Markdown template:

## 🔍 Code Review Report — [Branch/Task] ### ✅ Positives - [Design strengths, modularity, or good practices applied] ### 🔴 Blocking Findings (Require Mandatory Fix) - **[File:Line]** — [Description of security, build, or severe logic issue] ### 🟡 Suggested Improvements (Do Not Block Merge) - **[File:Line]** — [Readability recommendation, minor optimization, or future refactor] ### 🔵 Technical Notes / Debt - [Contextual notes on architecture or long-term behavior] ### 🛠️ FINAL VERDICT [APPROVED / REQUEST CHANGES]