What problem does it solve? Operations that run amid crashes, restarts, and retries often leave partial state behind, causing every restart to behave differently and turning recovery into a debugging session. This Skill provides a design discipline for making state-mutating operations converge to the same end state no matter how many times they run or where they start from. ## Core Features & Use Cases - Convergent Startup Pattern: Scan for existing state, clean stale artifacts, and adopt live sessions instead of assuming a fresh start. - Self-Healing Mechanisms: Apply content-based cleanup, PID-based stale lock detection, and idempotent scheduling that respawns failed work cleanly. - Idempotency Test: A three-question checklist (run twice, crash at any point, converge to same end state) to audit any operation. - Use Case: When writing a CLI command that creates worktrees and branches, apply this Skill to ensure re-running the command after a mid-execution crash reconciles leftover state instead of failing or duplicating resources. ## Quick Start Ask the AI to review your command or processing loop using the make-operations-idempotent principle and identify where partial state could break re-execution.