How does the Code Review skill handle large pull requests?

The skill automatically calibrates its review depth based on the changeset size. For very large PRs (50+ files), it prioritizes critical areas like security and core logic while providing a warning to ensure the most important changes are reviewed first.

What dimensions does the AI check during a review?

It evaluates code across eight specific dimensions: Correctness, Scope Fit, Design, Testing, Readability, Security, Performance, and Documentation.

Can this skill provide a final verdict on a PR?

Yes, it uses a weighted decision matrix to synthesize findings into a clear verdict: Approve, Request Changes, or Comment, based on the severity of the issues found.

Does it integrate with GitHub?

Yes, the skill is designed to work with the GitHub CLI (gh) to checkout pull requests and gather context from metadata and linked issues.

Automated Code Review

Name: Automated Code Review
Author: btimothy-har

bybtimothy-har

0•

Collaboration Tools

Conducts multi-dimensional, structured code reviews with calibrated severity levels and actionable feedback.

This skill transforms ad-hoc code reviews into a systematic evaluation process by assessing changes across eight key dimensions including security, performance, and readability. It utilizes specialized agents for context gathering and deep-dive analysis, providing a synthesized verdict ranging from approval to requested changes based on a weighted decision matrix. Ideal for maintaining high code quality standards while adapting review depth to the size and complexity of each pull request.

Key Features

01Calibrated severity levels from blockers to nitpicks for clear prioritization.

02Multi-dimensional analysis across 8 key quality areas including security and performance.

030 GitHub stars

04Structured feedback format with specific code suggestions and improvement examples.

05Size-based calibration that adjusts review depth for small to very large changesets.

06Automated context gathering using PR metadata, linked issues, and author intent.

Use Cases

01Automated first-pass reviews for pull requests to identify common bugs and design flaws.

02Security and performance auditing for critical code changes before production deployment.

03Standardizing code quality across teams by providing consistent, objective feedback.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add btimothy-har/basecamp code-review

For use in Claude.ai and ChatGPT