audit-engineering-productivity

name	audit-engineering-productivity
description	Run a single-session engineering productivity audit on the codebase
supports_parallel	true
fallback_available	true
estimated_time_parallel	15 min
estimated_time_sequential	45 min

Single-Session Engineering Productivity Audit

Purpose

Evaluates developer experience (DX), debugging capabilities, and offline support infrastructure. Identifies friction points in development workflows and provides actionable recommendations for improvement.

Execution Mode Selection

Condition	Mode	Time
Task tool available + no context pressure	Parallel	~15 min
Task tool unavailable	Sequential	~45 min
Context running low (<20% remaining)	Sequential	~45 min
User requests sequential	Sequential	~45 min

---

Section A: Parallel Architecture (3 Agents)

When to use: Task tool available, sufficient context budget

Step 0: Load False Positives

# Read false positives to avoid re-flagging known issues
cat docs/technical-debt/FALSE_POSITIVES.jsonl 2>/dev/null

Agent 1: dx-golden-path-auditor

Focus Areas:

• Developer onboarding friction
• npm scripts completeness
• Setup automation (doctor scripts, env validation)
• Development workflow ergonomics

Files:

• package.json (scripts section)
• scripts/ directory
• .env.example, .env.local
• README.md, DEVELOPMENT.md

Checks:

# Count npm scripts
grep -c '"[^"]*":' package.json

# Check for setup/doctor scripts
ls scripts/*.js 2>/dev/null | grep -E 'setup|doctor|verify'

# Check for dev:offline script
grep -c "dev:offline" package.json

Agent 2: debugging-ergonomics-auditor

Focus Areas:

• Structured logging vs console.log ratio
• Correlation ID support for request tracing
• Error message quality and user guidance
• Sentry integration completeness

Files:

• lib/logger.ts
• *.ts *.tsx files (console.log patterns)
• src/**/*.ts (backend/service logging)
• src/**/error.tsx (error boundaries)

Checks:

# Console.log to logger ratio
echo "Console.log calls:"
grep -r "console.log" --include="*.ts" --include="*.tsx" src/ lib/ components/ | wc -l
echo "Logger calls:"
grep -r "logger\." --include="*.ts" --include="*.tsx" src/ lib/ components/ | wc -l

# Correlation ID patterns
grep -r "correlationId\|correlation-id\|x-request-id" --include="*.ts" .

Agent 3: offline-support-auditor

Focus Areas:

• Offline persistence configuration (IndexedDB, localStorage, cache API)
• Service worker presence and functionality
• Offline write queue implementation
• Network status indicators and user feedback

Files:

• lib/persistence.ts or equivalent (if exists)
• public/sw.js (if exists)
• components/status/offline-indicator.tsx (if exists)
• lib/offline-queue.ts (if exists)

Checks:

# Offline persistence patterns
grep -r "enableIndexedDbPersistence\|enablePersistence\|IndexedDB\|idb\|Dexie" lib/ src/ 2>/dev/null || echo "No offline persistence found"

# Service worker
ls -la public/sw.js 2>/dev/null || echo "No service worker"

# LocalStorage vs IndexedDB usage
grep -r "localStorage" --include="*.ts" --include="*.tsx" | wc -l
grep -r "indexedDB\|IDB\|Dexie" --include="*.ts" --include="*.tsx" | wc -l

Parallel Execution Command

Invoke all 3 agents in a SINGLE Task message:

Task 1: dx-golden-path-auditor agent - audit setup, scripts, onboarding Task 2:
debugging-ergonomics-auditor agent - audit logging, error handling Task 3:
offline-support-auditor agent - audit persistence, service workers

Each agent prompt MUST end with:

CRITICAL RETURN PROTOCOL:

- Write findings to the specified output file using Write tool or Bash
- Return ONLY: `COMPLETE: [agent-id] wrote N findings to [output-path]`
- Do NOT return full findings content — orchestrator checks completion via file

Dependency constraints: All 3 agents are independent -- no ordering required. Each audits a separate domain and writes to separate output sections.

---

Section B: Sequential Fallback

When to use: Task tool unavailable, context pressure, or user request

Step 1: Golden Path & DX (10 min)

1. Review package.json scripts section
2. Check for setup/doctor automation scripts
3. Verify .env.example completeness
4. Assess README setup instructions
5. Identify missing "golden path" scripts

Step 2: Debugging Ergonomics (10 min)

1. Calculate console.log vs logger ratio
2. Search for correlation ID patterns
3. Review error boundary implementations
4. Check Sentry integration in logger
5. Assess error message quality

Step 3: Offline Support (10 min)

1. Check offline persistence configuration
2. Look for service worker implementation
3. Review offline indicator component
4. Search for write queue patterns
5. Assess localStorage vs IndexedDB usage

---

Output Requirements

Report File

Create docs/audits/single-session/engineering-productivity/audit-report.md:

# Engineering Productivity Audit - {DATE}

## Baselines

| Metric                   | Value |
| ------------------------ | ----- |
| npm scripts count        | X     |
| Console.log calls        | X     |
| Logger calls             | X     |
| Structured logging ratio | X:Y   |
| LocalStorage usages      | X     |
| IndexedDB usage          | X     |
| Service worker           | Y/N   |
| Offline persistence      | Y/N   |

## Findings Summary

| Severity | Count | Category       |
| -------- | ----- | -------------- |
| S0       | X     | -              |
| S1       | X     | Category names |
| S2       | X     | Category names |
| S3       | X     | Category names |

## Detailed Findings

### 1. Golden Path & DX

[Findings here]

### 2. Debugging Ergonomics

[Findings here]

### 3. Offline Support

[Findings here]

## Quick Wins (E0-E1)

[Prioritized list of low-effort fixes]

## Recommendations

### Immediate (E0-E1)

[List]

### Short-term (E2)

[List]

### Long-term (E3)

[List]

JSONL File

Create docs/audits/single-session/engineering-productivity/audit-findings.jsonl:

CRITICAL - Use this exact schema:

{
  "category": "engineering-productivity",
  "title": "Short specific title",
  "fingerprint": "engineering-productivity::primary_file::identifier",
  "severity": "S0|S1|S2|S3",
  "effort": "E0|E1|E2|E3",
  "confidence": 75,
  "files": ["array/of/file/paths.ts"],
  "why_it_matters": "1-3 sentences explaining impact",
  "suggested_fix": "Concrete remediation direction",
  "acceptance_tests": ["Array of verification steps"],
  "evidence": ["code snippet", "grep output", "measurement data"]
}

---

Category Definitions

Category	Description
GoldenPath	Setup, scripts, onboarding, dev workflow
Debugging	Logging, tracing, error handling, messages
Offline	Persistence, service workers, queues, sync

---

Severity Scale

Level	Name	Definition
S0	Critical	Data loss risk, production breaking
S1	High	Significant dev friction, missing core
S2	Medium	Maintainability drag, inconsistency
S3	Low	Polish, minor improvements, nice-to-have

---

Effort Scale

Level	Name	Definition
E0	Minutes	Quick fix, trivial change
E1	Hours	Single-session work
E2	Days	1-3 days or staged PR
E3	Weeks	Multi-PR, multi-week effort

---

Prior Audit Reference

Check docs/audits/single-session/engineering-productivity/ for prior findings and track carryover issues.

---

Standard Audit Procedures

Read CANON/templates/AUDIT_TEMPLATE.md for: Evidence Requirements, Dual-Pass Verification, Cross-Reference Validation, JSONL Output Format, Context Recovery, Post-Audit Validation, MASTER_DEBT Cross-Reference, Interactive Review, TDMS Intake & Commit, Documentation References, Agent Return Protocol, and Honesty Guardrails.

Skill-specific TDMS intake:

node scripts/debt/intake-audit.js \
  docs/audits/single-session/engineering-productivity/audit-findings.jsonl \
  --source "audit-engineering-productivity-$(date +%Y-%m-%d)"

---

When to Use

• User explicitly invokes /audit-engineering-productivity
• Engineering productivity or DX audit is needed
• Called as part of /audit-comprehensive (Stage 2)

When NOT to Use

• When you only need to audit code quality — use /audit-code instead
• When you need CI/CD pipeline health — use /audit-process instead
• When you need a comprehensive multi-domain audit — use /audit-comprehensive

---

Related Skills

• /audit-comprehensive - Run all domain audits
• /audit-process - CI/CD and automation audit (overlaps with DX)
• /audit-code - Code quality audit

---

Version History

Version	Date	Description
1.1	2026-02-23	Add mandatory MASTER_DEBT cross-reference step before interactive review
1.0	2026-02-03	Initial skill creation