error-recovery-and-resilience_skill

Overview

This skill teaches how to engineer resilient UI systems using layered error boundaries, retry policies, and fallback orchestration. It focuses on isolating failures so user-facing regions can recover independently without a full app reload. The guidance covers telemetry integration and chaos-testing strategies to validate improvements.

How this skill works

The skill inspects UI component hierarchies and recommends placement of nested error boundaries to contain failures. It defines retry policies (exponential backoff with jitter) and coordinates fallback UIs and automated recovery flows. Telemetry hooks record boundary activations, retries, and final outcomes to drive further tuning and alerting.

When to use it

• When single-component failures must not crash the entire application
• When network-dependent UI regions need automated retry and graceful degradation
• When you need observable recovery paths for incident analysis
• When introducing features that should tolerate partial outages
• When validating resilience improvements with chaos-testing

Best practices

• Place fine-grained error boundaries around independent UI regions, and broader boundaries at layout-level
• Combine exponential backoff with jitter to avoid synchronized retry storms
• Expose clear, minimal fallback UIs that offer retry or safe alternatives
• Emit structured telemetry for boundary triggers, retry attempts, and user actions
• Run targeted chaos tests and use results to refine containment and timeout thresholds

Example use cases

• Recovering a comments widget while keeping the rest of the page interactive
• Retrying a payment method fetch with exponential backoff and showing a cached fallback
• Orchestrating a staged fallback: lightweight view, then cached data, then retry option
• Measuring user impact by correlating boundary activations with session metrics
• Validating error containment via simulated downstream service failures