What is the review-fix-report protocol?

It is a tiered strategy where the agent reviews code changes, automatically fixes minor issues (under 20 lines), proposes options for moderate logic changes, and escalates architectural concerns.

Does it support automated testing frameworks?

Yes, it is pre-configured to work with pytest for targeted and full regression testing to verify all fixes and changes.

What inputs does the Review-Build skill accept?

You can provide a branch name, a specific commit range (SHA..SHA), a pasted build summary, or a Pull Request number.

How does this skill handle existing bugs in the codebase?

The skill runs a baseline check on the base commit before analyzing the new build, ensuring it only flags or fixes issues introduced by the latest changes.

Build Review & Repair Protocol

Name: Build Review & Repair Protocol
Author: marcusglee11

bymarcusglee11

0•

安全与测试

Performs comprehensive code reviews and automatically repairs obvious bugs using a tiered review-fix-report protocol.

The Review-Build skill implements a sophisticated 'review-fix-report' protocol designed to streamline the quality assurance process for new code contributions. By analyzing specific branches or commit ranges, it establishes a baseline for test failures to distinguish between pre-existing issues and those introduced in the current build. It categorizes findings into severity tiers, proactively fixing logic errors and style inconsistencies in-place while escalating larger architectural concerns to the user. This skill is essential for maintaining code integrity in multi-agent environments and ensuring every pull request meets high standards for correctness, safety, and testing coverage.

主要功能

01Baseline regression testing to identify pre-existing vs. new failures

02Standardized inter-agent reporting for consistent communication

03Comprehensive git diff analysis across branches, commits, or PRs

04Automated in-place repairs for straightforward bugs and pattern inconsistencies

05Tiered issue classification for critical, moderate, and low-priority findings

060 GitHub stars

使用场景

01Standardizing build reports and regression testing before merging to main

02Automating the review of code generated by other AI agents or team members

03Ensuring new features meet codebase-specific safety and consistency patterns

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add marcusglee11/lifeos review-build

For use in Claude.ai and ChatGPT

Download Skill