Repair Event Deduplication TTL Bugs

The problem

Event deduplication and TTL policies often fall out of sync across distributed repositories, leading to race conditions, duplicate side effects, or inconsistent state ownership. Manually debugging these failures is difficult due to non-deterministic concurrency issues and brittle retry logic.

What it does

• Identifies misaligned TTL and replay-window boundaries across repository components.
• Enforces atomic side-effect transitions to prevent partial failures during event processing.
• Maps event identity and producer scope to establish clear trust boundaries.
• Generates deterministic local fixtures and regression tests to simulate concurrency and out-of-order duplicates.
• Fixes crash-recovery and retry behavior without weakening existing event contracts.

Why this beats prompting it yourself

Generic prompts often suggest broad retries or thread sleeps that mask underlying correctness defects. This skill systematically audits state ownership and atomicity, ensuring repairs respect your existing retention policies and ordering guarantees without introducing security regressions.

Use cases

• Fixing idempotency bugs where the same event triggers multiple side effects.
• Aligning event retention periods with deduplication storage markers.
• Resolving race conditions in concurrent event processing pipelines.
• Implementing redacted observability for event lifecycle tracking.

Known limitations

Requires repository access via Read/Edit/Bash tools. Does not interact with live production environments, deployed state, or real customer secrets.