Debugging And Error Recovery

Debugging and Error Recovery (TypeScript API)

Overview

Use a controlled workflow to fix failures without expanding scope or introducing regressions.

When to Use

• Failing tests or runtime errors
• Behavior deviating from requirements
• Regression reports after recent changes

When NOT to Use

• Greenfield implementation without a specific failure signal

Inputs

• Error output and reproduction steps
• Relevant code paths and recent changes
• Existing tests around failing behavior

Process

1. Reproduce issue reliably.
2. Isolate failing boundary.
3. Add/adjust failing test for issue.
4. Implement minimal targeted fix.
5. Verify fix and regression safety.

Verification Checklist

• Reproduction exists before fix
• Root cause identified clearly
• Fix is minimal and scoped
• Related tests pass post-fix

Failure Modes to Avoid

• Guess-based fixes without reproduction
• Broad refactors during incident response
• Closing issue without regression checks

Output Requirements

• Use summarize for final output.
• Include root cause, fix summary, and verification evidence.