How does the skill handle UI-specific bugs?

It utilizes the agent-browser CLI to visually navigate applications and verify reported UI issues through automated browser control.

Can this skill identify if an issue is a user error rather than a bug?

Yes, part of the validation methodology involves checking the codebase's intended behavior against documentation and tests to classify issues as 'User Error' or 'Not a Bug' when appropriate.

What kind of output does the Bug Reproduction Validator provide?

It generates a structured report containing the reproduction status, steps taken, findings, root cause analysis, evidence (logs/code), and a severity assessment.

Does it work with specific programming frameworks?

While framework-agnostic, it includes specialized logic for modern environments like Rails, including checking development logs and database constraints.

How many times does it attempt to reproduce an issue?

The validator is designed to run reproduction steps at least twice to ensure consistency and eliminate intermittent or fluke results.

Bug Reproduction Validator

Name: Bug Reproduction Validator
Author: gvkhosla

bygvkhosla

•

安全与测试

Systematically validates and reproduces reported software issues to confirm genuine bugs and identify root causes.

The Bug Reproduction Validator is a specialized agent designed to streamline the debugging process by objectively verifying reported issues. It extracts critical information from bug reports, sets up minimal test environments, and executes reproduction steps methodically to distinguish between actual software defects, environment-specific issues, and user errors. By providing structured reports with severity assessments and root cause identification, it ensures developers focus their time on confirmed bugs with clear evidence.

主要功能

01Detailed structured reporting with evidence, root cause, and severity assessment

02Minimal test case setup and execution for consistent verification

03Comprehensive bug classification system (Confirmed, Cannot Reproduce, User Error, etc.)

04Cross-layer validation including UI (browser-based) and backend (logs/DB) analysis

05Automated extraction of reproduction steps and expected behaviors from reports

0635 GitHub stars

使用场景

01Verifying user-reported issues to confirm they are reproducible bugs before assigning resources

02Investigating complex edge cases involving specific data states or database constraints

03Filtering out environmental issues and user misunderstandings from the engineering backlog

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add gvkhosla/compound-engineering-pi bug-reproduction-validator

For use in Claude.ai and ChatGPT

Download Skill