skills.rest
susamn
susamnCOMMUNITY
Explore all skills by this author →
GitHub

pr-review

Compile GitHub PR metadata and diffs into structured LLM code review prompts.

Updated May 28, 2026
Search Tags:#github#code-quality#code-review#pull-request#pr-review#llm-prompt#diff-chunking
One-click install
npx skills add https://github.com/susamn/skills --skill pr-review-susamn
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: pr-review
Source: https://github.com/susamn/skills/tree/main/pr-review
Command: npx skills add https://github.com/susamn/skills --skill pr-review-susamn

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires gh, jq, and includes scripts (resource) components.

What problem does it solve?

Manually compiling GitHub pull request metadata, diffs, and context into structured prompts for LLM code reviews is time-consuming and error-prone, especially for large PRs with diffs that exceed LLM context limits. This skill automates the entire process end-to-end.

Core Features & Use Cases

  • Automated PR Data Fetching: Pulls full PR metadata, commit history, changed files, and diffs directly from GitHub via the GitHub CLI.
  • Context-Aware Prompt Generation: Injects optional user story or ticket context into prompts for more relevant, targeted code reviews.
  • Diff Chunking for LLM Limits: Automatically splits large PR diffs into approximately 100KB chunks to fit within standard LLM context windows, with clear chunk metadata for each generated prompt.
  • Use Case: A software developer needs to conduct a thorough code review of a 300KB PR diff. This skill generates multiple separate, ready-to-use LLM prompts that cover 9 key code review dimensions, saving hours of manual copy-pasting and formatting work.

Quick Start

Use the pr-review skill to generate a structured LLM code review prompt for the GitHub pull request at https://github.com/your-org/your-repo/pull/123, and include the user story from the file feature-456.md for additional context.

Frequently Asked Questions about pr-review

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

FAQPage Schema
How do I generate LLM code review prompts for large GitHub pull requests?▼

To generate LLM code review prompts for large GitHub pull requests, this skill fetches PR metadata and diffs via the GitHub CLI, automatically splitting large diffs into context-sized chunks to fit within standard LLM context limits.

What is diff chunking and how does it handle pull requests that exceed LLM context limits?▼

Diff chunking handles pull requests exceeding LLM context limits by automatically splitting large PR diffs into approximately 100KB chunks, generating separate structured LLM-ready prompts with clear chunk metadata for each segment.

How do I add user story context to an automated LLM code review prompt?▼

To add user story context to an automated LLM code review prompt, the skill injects optional user story or ticket context directly from a specified file into the generated prompt, ensuring more relevant and targeted code reviews.

Do I need the GitHub CLI installed to automate pull request code reviews?▼

Yes, you need the GitHub CLI installed to automate pull request code reviews, as the skill leverages the GitHub CLI to fetch full PR metadata, commit history, changed files, and diffs directly from your repository.

What code review dimensions are covered by automated LLM prompt generation?▼

Automated LLM prompt generation covers 9 standard code review dimensions, explicitly including correctness, security, and performance, ensuring a thorough and structured evaluation of the pull request changes.

Related Skills

Explore More Software Engineering Skills→
githubnextgithubnext
4.9k

github-pr-query

Query GitHub pull requests with optional jq filtering and JSON output.

Official
Intermediate
YomiamyYomiamy
16

receiving-code-review

Manage code review feedback with structured verification and questioning.

Community
Intermediate
TabTabGoTabTabGo

expo-cicd-workflows

Generate, edit, and validate Expo EAS workflow YAML files against schema standards.

Official
Basic
framerslabframerslab

code-safety

Scan LLM-generated code for injection, XSS, and security vulnerabilities.

Official
Intermediate