Does this work with my existing test suite?

Yes, it is designed to integrate with standard test runners like pnpm or jest and requires your project's test infrastructure to be in place.

How does this skill improve the bug fixing process?

It enforces a systematic Test-Driven Development (TDD) process that requires a failing reproduction test before any code is changed, ensuring the fix is verified and regressions are prevented.

What documentation does the skill generate?

It generates an 'Evidence Triplet' consisting of a reproduction proof in tests.md, a contract snapshot of types/schemas, and a review.json file containing the full findings.

Can I use this skill for high-stakes production incidents?

Yes. The skill includes a 'Consultative' mode where the AI performs the analysis and proposes options while the human developer retains full decision-making power for the fix.

Bug Fix Workflow

Name: Bug Fix Workflow
Author: jscraik

byjscraik

セキュリティとテスト

Implements a disciplined, TDD-driven approach to identifying, reproducing, and resolving software defects with guaranteed verification.

概要

The Bug Fix skill transforms how Claude handles software defects by enforcing a rigorous Test-Driven Development (TDD) cycle. Instead of applying speculative patches, it guides the AI to isolate the failure, document a testable hypothesis, and write a failing reproduction test before implementing the most minimal fix possible. This workflow ensures that every resolution is backed by evidence, covers edge cases, and includes a full post-mortem documentation suite to prevent future regressions.

主な機能

Layered testing execution to prevent regression across the codebase
Strict TDD workflow: Reproduction, Hypothesis, Test, Fix, Verify
0 GitHub stars
Automated reproduction case isolation and documentation
Configurable AI participation modes from Delegated to Consultative
Comprehensive evidence generation for root-cause analysis and post-mortems

ユースケース

Resolving complex logic errors that require specific environment reproduction
Handling high-priority production incidents with structured root-cause documentation
Standardizing bug resolution patterns across large development teams

概要

主な機能

Layered testing execution to prevent regression across the codebase
Strict TDD workflow: Reproduction, Hypothesis, Test, Fix, Verify
0 GitHub stars
Automated reproduction case isolation and documentation
Configurable AI participation modes from Delegated to Consultative
Comprehensive evidence generation for root-cause analysis and post-mortems

ユースケース

Resolving complex logic errors that require specific environment reproduction
Handling high-priority production incidents with structured root-cause documentation
Standardizing bug resolution patterns across large development teams