What problem does it solve?
This Skill helps you find bugs caused by inconsistent state across lifecycles, counters, queues, events, retries, and external systems. It is designed for situations where something was acquired but never released, counted twice, lost in transit, or allowed to drift away from the authoritative source of truth.
Core Features & Use Cases
- Ledger-style state auditing: Maps every important transition to its required counterpart, such as acquire/release, enqueue/dequeue, emit/handle, or charge/refund.
- Defect detection framework: Systematically checks for leaks, orphans, duplicates, drift, and impossible states, then turns findings into concrete invariants and reconciliation rules.
- Cross-boundary reconciliation: Useful for reviewing distributed flows involving databases, caches, queues, payment providers, search indexes, or replicated systems where retries and partial failures can silently corrupt state.
- Use Case: If your service shows mismatched queue counts, duplicated payments, lingering sessions, or resources that never clean up after cancellation or crashes, this Skill gives you a structured review process to identify the broken counterpart and define automatic safeguards.
Quick Start
Ask the ledger-consistency skill to audit a specific lifecycle such as pending jobs, open connections, payment state, or event delivery and identify missing counterparts, drift, and invariants to enforce.