sentry-analysis

name	sentry-analysis
description	Analyze and fix production errors from Sentry crash reports. Use when investigating AppHang, ANR, crashes, or production errors. Includes complete workflow from JSON analysis to bug fix implementation with evidence-based methodology. Triggers on sentry, crash, AppHang, ANR, error analysis, production error, bug analysis, crash report, freeze, hang, not responding, stacktrace, breadcrumbs, exception.
allowed-tools	Read, Grep, Glob, Write, Edit, Bash

Sentry Error Analysis & Fixes

Complete workflow for analyzing and fixing production errors from Sentry crash reports.

Workflow Overview

1. Obtain Sentry JSON log
   ↓
2. Analyze error
   ↓
3. Identify root cause
   ↓
4. Generate bug analysis log
   ↓
🚨 WAIT FOR USER CONFIRMATION 🚨
   ↓
5. Implement fix (only after approval)
   ↓
6. Test & verify
   ↓
7. Create PR

Critical Requirements

MUST follow these rules:

1. ✅ Always create a bug analysis log in node_modules/.cache/bugs/ before implementing fixes
2. 🚨 MUST wait for user confirmation before starting any code changes
3. ✅ Bug analysis must be complete with all sections filled
4. ✅ Use evidence-based methodology (环环相扣，逐步递进)

Quick Reference

Common Error Types

Type	Description	Common Causes
AppHang	iOS app frozen >5s	Too many concurrent requests, main thread blocking
ANR	Android Not Responding	Heavy operations on main thread, deadlocks
Crash	App terminated	Null pointer, memory issues, unhandled exceptions
Exception	Handled error	Network failures, validation errors, state issues

Analysis Methodology

Use 6 types of proof to establish causation:

1. Stack Trace Evidence - Error location in code
2. Breadcrumbs Evidence - User actions leading to error
3. Code Logic Evidence - Why the code causes the issue
4. Timing Evidence - When and how often it occurs
5. Device/Platform Evidence - Affected platforms/devices
6. Fix Verification - Testing confirms fix works

Common Fix Patterns

// Pattern 1: Concurrent request control
async function executeBatched<T>(
  tasks: Array<() => Promise<T>>,
  concurrency = 3,
): Promise<Array<PromiseSettledResult<T>>> {
  const results: Array<PromiseSettledResult<T>> = [];
  for (let i = 0; i < tasks.length; i += concurrency) {
    const batch = tasks.slice(i, i + concurrency);
    const batchResults = await Promise.allSettled(
      batch.map((task) => task()),
    );
    results.push(...batchResults);
  }
  return results;
}

// Pattern 2: Main thread offloading (React Native)
import { InteractionManager } from 'react-native';

InteractionManager.runAfterInteractions(() => {
  // Heavy operation here
});

// Pattern 3: Error boundary
<ErrorBoundary fallback={<ErrorFallback />}>
  <Component />

Detailed Guide

For comprehensive Sentry error analysis workflow, see fix-sentry-errors.md.

Topics covered:

• Obtaining Sentry JSON logs
• Python-based quick analysis
• Bug analysis log template
• 6 types of proof methodology
• Root cause identification
• Common fix patterns (AppHang, ANR, Crashes)
• Real-world case studies
• Testing and verification
• PR creation workflow

Key Files

Purpose	Location
Bug analysis logs	node_modules/.cache/bugs/
Sentry config	packages/shared/src/modules/sentry/
Error boundaries	packages/kit/src/components/ErrorBoundary/

When to Use This Skill

• Analyzing iOS AppHang errors (5+ second freezes)
• Fixing Android ANR (Application Not Responding)
• Investigating crash reports with stack traces
• Understanding user actions before crashes (breadcrumbs)
• Creating evidence-based bug analysis reports
• Implementing fixes for production errors

Related Skills

• /performance - Performance optimization patterns
• /error-handling - Error handling best practices
• /sentry - Sentry configuration and filtering
• /code-quality - Lint fixes and code quality