audit

Audit Command

Scan a codebase for existing observability instrumentation and identify gaps.

Usage

/audit [path]

Arguments:

• path: Optional. Directory to scan (defaults to current directory)
  • Non-existent path: Error with "Path not found: [path]"
  • Non-directory: Error with "Path must be a directory: [path]"

Examples:

/audit
/audit ./src
/audit ~/projects/my-app

---

Workflow

Step 1: Launch Codebase Analyzer

Invoke the codebase-analyzer agent to:

1. Detect platform (iOS/Android/React Native)
2. Identify architecture patterns
3. Find existing telemetry SDKs
4. Map instrumentation coverage

Step 2: Load Reference Context

Read:

• references/instrumentation-patterns.md - Instrumentation checklist
• references/platforms/*.md - Vendor-specific patterns

Step 3: Evaluate Coverage

Check each instrumentation area against the codebase:

Area	What to Find	Priority
SDK Init	SDK setup in AppDelegate/Application	P0
Crash Reporting	Error capture, crash handlers	P0
User Context	setUser, user ID setting	P0
Symbolication	dSYM upload, ProGuard config	P0
App Start	Launch timing, startup spans	P1
Screen Load	TTI tracking, screen spans	P1
Network	Request interceptors, API tracing	P1
Breadcrumbs	Navigation/action logging	P1
Database	Query tracing	P2
Custom Spans	Business flow instrumentation	P2

Step 4: Identify Anti-Patterns

Flag issues like:

• PII in breadcrumbs/logs
• Synchronous telemetry on main thread
• Missing symbolication config
• Over-instrumentation (battery impact)
• Inconsistent span naming

Step 5: Generate Report

Output format:

## Telemetry Audit Report

### Platform
- **Type:** [iOS/Android/React Native]
- **Language:** [Swift/Kotlin/TypeScript]

### Existing SDKs

| SDK | Version | Location | Status |
|-----|---------|----------|--------|
| Sentry | 8.0.0 | Podfile:23 | Active |
| Firebase | 10.0.0 | Podfile:25 | Crashlytics only |

### Coverage Assessment

| Area | Status | Details |
|------|--------|---------|
| Crash Reporting | ✅ Covered | Sentry SDK initialized in AppDelegate |
| Error Tracking | ⚠️ Partial | Only in network layer |
| App Start | ❌ Missing | No launch timing |
| Screen Load | ⚠️ Partial | 3/12 screens tracked |
| Network | ✅ Covered | URLSession instrumented |
| Breadcrumbs | ❌ Missing | No breadcrumb logging |
| User Context | ✅ Covered | User ID set on login |
| Symbolication | ⚠️ Partial | dSYM upload configured, but not verified |

### Coverage Score: 62%

Foundation: ████████░░ 80% Performance: ████░░░░░░ 40% Context: ██░░░░░░░░ 20% Advanced: ░░░░░░░░░░ 0%

### Gaps Identified

| Gap | Priority | Impact | Effort |
|-----|----------|--------|--------|
| No app start tracking | P1 | Can't measure launch perf | Low |
| No screen load TTI | P1 | Can't identify slow screens | Medium |
| No breadcrumbs | P1 | Poor crash debugging | Low |
| Missing database spans | P2 | Can't trace slow queries | Medium |

### Anti-Patterns Found

1. **[Issue]** - [Location] - [Recommendation]

### Recommendations

#### Quick Wins (This Week)
1. [Low effort, high impact items]

#### Medium Term (This Month)
1. [Moderate effort items]

#### Nice to Have
1. [Lower priority items]