What problem does it solve?
Skills and automation workflows become unreliable when failing scripts or unclear instructions are worked around instead of fixed, causing silent divergence and recurring bugs. This Skill enforces a fix-and-retry discipline so that failures are diagnosed, corrected, and re-run, preserving deterministic workflows and auditability.
Core Features & Use Cases
- Fix-first policy: Require explicit identification of the failing skill step, a proposed corrective change, and re-running the original skill instead of manual workarounds.
- Failure reporting guidance: Standardizes how to report syntax errors, runtime errors, missing paths, or incorrect outputs so fixes are actionable.
- Use Case: In a CI pipeline or developer workflow where a data-gathering script fails, follow the skill-repair steps to fix the script, validate output, and prevent ad-hoc bypasses that hide the bug.
Quick Start
When a skill script errors, describe the exact failure, propose a specific fix, apply the fix with confirmation if unsure, and re-run the original skill step.