peter-bugfix-loop

What This Skill Does

The peter-bugfix-loop is a structured engineering workflow agent designed to formalize the software bug resolution process within the OpenClaw environment. Rather than allowing ad-hoc fixes, this skill forces a disciplined 'reproduce-isolate-fix-verify' lifecycle. It acts as an automated technical assistant that ensures developers prioritize evidence-based debugging, preventing the common mistake of 'fixing' issues by trial and error without proper verification. The skill mandates the creation of an explicit Bug Card, ensures that a failing test case (Red) is established before any code modification occurs, and guides the user through root cause analysis via hypothesis testing. By integrating with existing Peter DevOps tools, it ensures that every fix is accompanied by relevant unit or integration tests and adheres to professional code review and CI gating standards, effectively narrowing the 'blast radius' of changes and improving long-term system stability.

Installation

To install this skill, run the following command in your terminal within your OpenClaw environment: clawhub install openclaw/skills/skills/chinasilva/peter-bugfix-loop

Use Cases

• Production Outages: Rapidly triaging and patching runtime errors by moving from incident report to a verified, test-covered fix.
• CI/CD Stabilization: Identifying whether failing pipeline tests are due to genuine regression or environment noise, and ensuring permanent automated coverage.
• Legacy Code Maintenance: Providing a structured, low-risk approach to debugging complex, undocumented codebases by enforcing minimal changes.
• Onboarding/Standardization: Teaching junior engineers a repeatable, professional standard for bug investigation and resolution.

Example Prompts

1. "I'm seeing 500 errors on the checkout endpoint. Please start a peter-bugfix-loop to track and resolve this."
2. "CI is failing on the login integration test. Can you use peter-bugfix-loop to help me reproduce and fix the underlying issue?"
3. "User reports that the export function hangs with large datasets. Let's run a bugfix-loop to find the root cause and patch it."

Tips & Limitations

• Tip: Always define a clear 'Acceptance Sentence' in the Bug Card stage; this is crucial for the AI to understand when the fix is considered complete.
• Tip: If you cannot immediately write an automated test, at least provide a clear shell script or manual reproduction steps to satisfy the '先红' (first-red) constraint.
• Limitation: The skill strictly prohibits 'code-first' fixes. If you skip the reproduction step, the skill will nudge you back to maintain safety.
• Limitation: It is designed for individual issue resolution; do not use it for large-scale feature development or massive architectural refactoring.