debugging-strategies
This Claude Code skill provides language-agnostic debugging techniques applicable across all technology stacks, including the scientific method for debugging, systematic approaches to reproducing and isolating bugs, stack trace analysis, and specific techniques like binary search and the wolf fence algorithm. Use this when troubleshooting issues, analyzing error messages, or establishing a structured debugging methodology for any programming language or framework.
git clone --depth 1 https://github.com/MadAppGang/claude-code /tmp/debugging-strategies && cp -r /tmp/debugging-strategies/plugins/dev/skills/core/debugging-strategies ~/.claude/skills/debugging-strategiesSKILL.md
# Universal Debugging Strategies
## Overview
Language-agnostic debugging techniques and strategies applicable across all technology stacks.
## Debugging Methodology
### The Scientific Method for Debugging
1. **Observe**: Gather information about the bug
2. **Hypothesize**: Form theories about the cause
3. **Predict**: What would confirm/refute each theory?
4. **Test**: Run experiments to validate
5. **Conclude**: Identify root cause and fix
### Systematic Approach
```
1. Reproduce the bug reliably
2. Isolate the failing code path
3. Trace backwards from the error
4. Identify the root cause
5. Verify the fix
6. Add tests to prevent regression
```
## Error Analysis
### Stack Trace Reading
Read stack traces **bottom to top**:
```
Error: Cannot read property 'name' of undefined
at formatUser (src/utils/format.ts:42) ← Error thrown here
at processUsers (src/services/user.ts:28) ← Called from here
at UserList.render (src/components/UserList.tsx:15)
at App.render (src/App.tsx:8) ← Entry point
```
**Focus on YOUR code** - skip framework/library internals initially.
### Error Categories
| Category | Examples | Common Causes |
|----------|----------|---------------|
| **Null/Undefined** | `Cannot read property 'x' of undefined` | Missing data, async timing |
| **Type Errors** | `x is not a function` | Wrong type, typo |
| **Logic Errors** | Wrong output, no error | Incorrect conditions |
| **Runtime Errors** | Out of bounds, division by zero | Invalid input |
| **Async Errors** | Unhandled promise rejection | Missing error handler |
| **Network Errors** | Timeout, connection refused | API down, wrong URL |
## Debugging Techniques
### Binary Search (Bisection)
When the bug is somewhere in a large codebase:
```
1. Find a known good state (commit, version)
2. Find current bad state
3. Test the midpoint
4. Narrow to half with bug
5. Repeat until found
```
**Git bisect**:
```bash
git bisect start
git bisect bad # Current commit is broken
git bisect good v1.0.0 # v1.0.0 was working
# Git checks out midpoint, test it
git bisect good/bad # Mark and continue
```
### Wolf Fence Algorithm
Split the code into sections and determine which section contains the bug:
```
┌─────────────────────────────────────┐
│ Code Section A │ ← Test: Bug here?
├─────────────────────────────────────┤
│ Code Section B │ ← Test: Bug here?
├─────────────────────────────────────┤
│ Code Section C │ ← Test: Bug here?
└─────────────────────────────────────┘
```
Add logging/breakpoints at section boundaries, narrow down.
### Rubber Duck Debugging
Explain the code line by line (to a rubber duck, colleague, or yourself):
1. Explain what the code SHOULD do
2. Explain what it ACTUALLY does
3. The discrepancy reveals the bug
### Change One Thing at a Time
When experimenting:
- Make ONE change
- Test
- Observe result
- Revert if no improvement
- Repeat
## Data Flow Tracing
### Backwards Tracing
Start at the error, trace backwards:
```
1. Error occurs at line 42: user.name is undefined
2. Where does `user` come from? Line 38: const user = getUser(id)
3. What does getUser return? Check function...
4. getUser queries database, returns undefined if not found
5. Root cause: No null check after getUser
```
### Forward Tracing
Start at input, trace forward:
```
1. User enters email: "test@example"
2. Form submits to /api/register
3. API validates email... passes (bug: missing TLD check)
4. Saves to database with invalid email
5. Later processes fail on invalid email
```
## Logging Strategies
### Strategic Log Placement
```typescript
function processOrder(order) {
console.log('processOrder START', { orderId: order.id });
try {
const validated = validateOrder(order);
console.log('Validation passed', { orderId: order.id });
const result = saveOrder(validated);
console.log('processOrder SUCCESS', { orderId: order.id, result });
return result;
} catch (error) {
console.error('processOrder FAILED', {
orderId: order.id,
error: error.message,
stack: error.stack
});
throw error;
}
}
```
### Log Levels
| Level | Use For | Example |
|-------|---------|---------|
| **ERROR** | Failures requiring attention | `Failed to save order: DB connection lost` |
| **WARN** | Potential issues | `Retry 2/3 for API call` |
| **INFO** | Significant events | `User logged in: user123` |
| **DEBUG** | Detailed diagnostics | `Validating email: test@example.com` |
| **TRACE** | Very verbose | `Entering function processOrder` |
### Structured Logging
```typescript
// BAD: Unstructured
console.log('User ' + userId + ' ordered ' + items.length + ' items');
// GOOD: Structured
logger.info('Order placed', {
userId,
itemCount: items.length,
total: order.total,
timestamp: new Date().toISOString()
});
```
## Breakpoint Strategies
### Types of Breakpoints
| Type | Use Case |
|------|----------|
| **Line** | Stop at specific line |
| **Conditional** | Stop only when condition is true |
| **Logpoint** | Log without stopping |
| **Exception** | Stop on thrown exception |
| **DOM** | Stop on DOM modification (browser) |
### Effective Breakpoint Placement
1. **Before the error** - See state leading to failure
2. **At decision points** - Check which branch executes
3. **At data boundaries** - API calls, DB queries
4. **In loops** - Check iteration values
## Common Bug Patterns
### Off-by-One Errors
```typescript
// BUG: Index out of bounds
for (let i = 0; i <= array.length; i++) {
console.log(array[i]); // Fails on last iteration
}
// FIX
for (let i = 0; i < array.length; i++) {
console.log(array[i]);
}
```
### Race Conditions
```typescript
// BUG: Race condition
async function getData() {
fetchData().then(data => { this.data = data; });
processData(this.data); // May run before fetch completes!
}
// FIX: Await the result|
|
|
Common agent patterns and templates for Claude Code. Use when implementing agents to follow proven patterns for Tasks integration, quality checks, and external model invocation via claudish CLI.
YAML frontmatter schemas for Claude Code agents and commands. Use when creating or validating agent/command files.
XML tag structure patterns for Claude Code agents and commands. Use when designing or implementing agents to ensure proper XML structure following Anthropic best practices.
YAML format for Claude Code agent definitions as alternative to markdown. Use when creating agents with YAML, converting markdown agents to YAML, or validating YAML agent schemas. Trigger keywords - "YAML agent", "agent YAML", "YAML format", "agent schema", "YAML definition", "convert to YAML".
Linear API patterns and examples for autopilot. Includes authentication, webhooks, issue CRUD, state transitions, file attachments, and comment handling.