bug-reproduction-validator_skill

Overview

This skill is a Bug Reproduction Specialist that verifies whether reported issues are genuine bugs by systematically reproducing, validating, and classifying reported behavior. It produces a concise, evidence-backed report that states reproduction status, findings, root cause (if found), severity, and recommended next steps. Use it when you need a defensible determination of whether to open a fix, gather more info, or close a report.

How this skill works

The agent extracts critical reproduction details (steps, expected vs actual behavior, environment, logs) and sets up a minimal test case. It runs the reproduction steps methodically, checks edge cases and environments, inspects relevant code and recent git changes, and collects logs or test outputs. After repeated attempts it classifies the issue and assembles a structured report with evidence and remediation guidance.

When to use it

• You receive a bug report and need a reproducible validation before triage.
• You must decide whether to assign engineering time for a fix.
• Reports include partial steps or inconsistent results across environments.
• You need a clear, auditable reproduction record for QA or stakeholders.
• You want to determine whether an issue is environment- or data-specific.

Best practices

• Start by extracting exact reproduction steps and required environment details.
• Create the minimal test fixture or dataset needed to reproduce the problem.
• Run reproduction attempts at least twice and try boundary/edge inputs.
• Check recent git history and existing tests to validate intended behavior.
• Attach logs, code snippets, and concrete commands or browser interactions as evidence.

Example use cases

• Validate a report that email processing fails for subjects with special characters by reproducing processing flow and checking logs.
• Confirm whether a summary view omission is an application bug or due to data-range filters.
• Distinguish between a user error and a backend validation failure in an API endpoint.
• Reproduce a UI rendering issue using a headless browser and verify against CSS/templating code.
• Diagnose intermittent failures to decide if they are environmental or code regressions.