error-retry-tracking

What problem does it solve?

Instrument error handling to understand failure patterns and recovery behavior across distributed components.

Core Features & Use Cases

• Error Classification: categorize transient vs permanent failures to guide retries and fallbacks.
• Retry & Fallback Telemetry: capture retry counts, delays, outcomes, and fallback transitions to optimize strategies.
• Framework Integration: supports decorators and patterns (e.g., with_retry) to instrument LLM/tool calls and API requests.
• Error Aggregation: summarize errors per session or provider to monitor health and regression.
• Guardrails & Anti-patterns: detect common pitfalls and improve resilience with precautionary checks.

Quick Start

Wrap your callables with the with_retry decorator to start collecting retry telemetry and error context.