extension-audit-methodology

name	extension-audit-methodology
description	Systematic 5-dimension audit framework for VS Code extensions — debug hygiene, dead code, performance, menu validation, dependency cleanup
lastReviewed	2026-04-30T00:00:00.000Z

Extension Audit Methodology

Find quality issues standard code review misses — before release, not after deployment.

---

Pattern Overview

A systematic 5-dimension audit framework for VS Code extensions to identify code quality issues beyond standard linting. Produces actionable recommendations with evidence-based prioritization.

Problem Solved: Standard code review catches logic bugs and style issues, but misses extension-specific quality problems like orphaned commands, blocking I/O in UI threads, debug noise in production, and leftover dependencies from removed features.

When to Use:

• Before major release (milestone versions)
• After significant refactoring
• When performance degradation suspected
• Post-feature-removal cleanup

Time Investment: 2-4 hours for comprehensive audit + 4-8 hours remediation

---

5-Dimension Audit Framework

1. Debug & Logging Audit

Objective: Identify console statements and categorize by necessity

Method:

# Scan for all console statements
Select-String -Path src/*.ts -Pattern "console\.(log|debug|warn|error)" -Recurse

Categorization:

Category	Keep?	Examples	Rationale
Enterprise compliance	✅	Audit logs, security events	Legal/regulatory requirement
User feedback	✅	TTS status, long-running ops	User-facing information
Critical errors	✅	Unrecoverable failures	Debugging user-reported issues
Setup verbosity	❌	"Loading CSS..."	Noise during normal operation
Migration debugging	❌	Version upgrade steps	Temporary development artifacts

---

2. Dead Code Detection

Commands Cross-Check:

1. Find registered commands in extension.ts
2. Find command contributions in package.json
3. Find command invocations in UI/views
4. Identify: broken (in UI, not registered), dead (registered, never invoked)

---

3. Performance Profiling

Blocking I/O Scan:

# Synchronous file operations (event loop blockers)
Select-String -Path src/*.ts -Pattern "fs\.(readFileSync|existsSync|readdirSync|writeFileSync)" -Recurse

Refactoring Pattern:

// BEFORE (blocking)
const exists = fs.existsSync(path);

// AFTER (async)
import * as fs from 'fs-extra';
const exists = await fs.pathExists(path);

---

4. Menu Validation

Test Matrix:

• Extract all commands from package.json contributions
• Verify each has a registered handler
• Test in VS Code: Command Palette, buttons, context menus

---

5. Dependency Hygiene

Unused Package Detection:

npm ls --depth=0
# For each dependency, verify it's imported
Select-String -Path src/*.ts -Pattern "from '@package-name'" -Recurse

---

Audit Report Template

# Extension Audit Report

**Date**: YYYY-MM-DD
**Version**: X.Y.Z

## Executive Summary

- **Console statements**: X remaining (Y legitimate, Z removable)
- **Dead code**: [count] commands/features
- **Performance**: [X blocking ops]
- **Menu validation**: Y/X working (Z% functional)
- **Dependencies**: W unused packages

**Code Quality Grade**: [A+/A/A-/B+/B]

---

## Recommendations

| Priority | Category | Issue | Action | Effort |
|----------|----------|-------|--------|--------|
| Critical | Dead Code | Broken command | Fix or remove | 30m |
| High | Performance | Blocking I/O | Async refactor | 2h |
| High | Logging | Removable logs | Delete | 1h |
| Medium | Dependencies | Unused package | Remove | 15m |

---

Success Metrics

Before Audit (typical mature extension):

• Console statements: 40-80
• Dead commands: 3-8
• Blocking sync I/O: 10-30 operations
• Code quality grade: B/B+

After Remediation:

• Console statements: 15-25 (legitimate only)
• Dead commands: 0
• Blocking sync I/O: 0 (all async)
• Code quality grade: A/A+

Improvement: ~50% console reduction, 100% dead code removal, measurable performance gain

---

Real-World Application: the AI assistant v5.7.1

Audit Results:

• Console statements: 46 found → 27 removed (18 legitimate kept)
• Dead code: 3 deprecated commands + 1 UI feature + 1 unused dependency
• Performance: 16 blocking sync operations in cognitiveDashboard.ts
• Menu validation: 39/41 working (95%)

Remediation Actions:

1. Removed 27 console statements
2. Deleted 3 deprecated Gist sync commands
3. Removed generateImageFromSelection UI from 3 locations
4. Refactored dashboard to async I/O (fs → fs-extra)
5. Removed @azure/msal-node dependency (unused)
6. Deleted 477-line obsolete MS Graph documentation

Results:

• Code quality: B+ → A+
• TypeScript errors: 0
• Extension size: 9.45 MB (optimized)
• Console cleanliness: 61 total removed across 2 releases

Effort: 8 hours audit + remediation → measurable quality improvement

---

Post-Audit Verification Checklist

• TypeScript compiles: npm run compile → exit 0, zero errors
• No orphaned imports: All import statements resolve
• Extension activates: Test .vsix installation
• Smoke test: 3 random commands execute correctly
• Performance: Dashboard/views render without lag
• Console clean: Only intentional logs during normal operation