What problem does it solve?
This skill provides principled guidance for configuring resilience controls (circuit breakers, timeouts, and retries) by explaining the rationale behind chosen values, rather than focusing on implementation details. It covers threshold derivation, ignoreExceptions traps, separation of timeout and circuit breaker responsibilities, and retry target selection for external API integration.
Core Features & Use Cases
- Guidance on sizing circuit breakers based on observed steady-state failure rates rather than arbitrary numbers.
- Clear criteria for when to retry, which exceptions to record or ignore, and how to separate timeout handling from breaker state.
- Applicability to API clients and asynchronous pipelines (e.g., Kafka) to reduce oscillation and lag through tuned thresholds and probe settings.
- Practical anti-patterns and best practices with concrete examples.
Quick Start
Provide sizing guidance for circuit breakers and timeouts based on observed failure rates to design a robust external API integration.