Env Doctor

Env Doctor Skill

When to activate

• Starting on a new machine or fresh clone and something isn't working
• After pulling changes and getting unexpected errors
• Onboarding a new team member and verifying their setup
• Before a deploy or CI run to catch environment mismatches
• "It works on my machine" debugging

When NOT to use

• Runtime application errors unrelated to environment setup
• Performance issues that aren't environment-caused
• Production debugging (use your observability stack)

Instructions

What env-doctor checks

1. Runtime versions

node --version    # vs .nvmrc or engines.node in package.json
python --version  # vs .python-version or pyproject.toml
ruby --version    # vs .ruby-version
go version        # vs go.mod
java -version     # vs .java-version or pom.xml

2. Required env vars Reads from:

• .env.example — vars with no default = required
• CLAUDE.md environment section
• README "Getting Started" section
• Direct process.env.X / os.environ["X"] references in code

3. Dependencies installed

# Node
node_modules/ exists? package-lock.json matches package.json?

# Python
.venv/ exists? pip freeze matches requirements.txt / pyproject.toml?

# Ruby
bundle check

# Go
go mod verify

4. Services reachable

# Database
pg_isready -h localhost -p 5432

# Redis
redis-cli ping

# Other services from .env DATABASE_URL, REDIS_URL, etc.

5. Port conflicts

# Check if the dev server port is already in use
lsof -i :3000
lsof -i :8000
lsof -i :5432

6. Required CLI tools

which git gh docker docker-compose terraform kubectl aws gcloud

Usage

Just invoke the skill and Claude will run the checks for your project type:

/env-doctor

Or target a specific area:

/env-doctor --check envvars
/env-doctor --check deps
/env-doctor --check services

Output format

Claude produces a report:

🏥 Environment Health Report
════════════════════════════

✅ Node.js: 22.16.0 (required: >=18, matches .nvmrc)
✅ npm: 10.8.0

⚠️  Missing env vars (2):
   - STRIPE_WEBHOOK_SECRET — required for billing to work locally
     Set it: copy from 1Password > Dev Secrets > Stripe
   - OPENAI_API_KEY — required for AI features
     Set it: https://platform.openai.com/api-keys

❌ Services not reachable (1):
   - PostgreSQL on localhost:5432 — connection refused
     Fix: docker compose up -d db

✅ Dependencies: node_modules present, lockfile matches
✅ No port conflicts on :3000

Summary: 1 blocker, 2 warnings
Run `docker compose up -d db` to fix the blocker.

Quick fixes Claude suggests

Adding env-doctor to your project

Add to your project's CLAUDE.md:

## Environment Setup
Required env vars (all must be set for local dev):
- DATABASE_URL — PostgreSQL connection string
- REDIS_URL — Redis connection string
- STRIPE_SECRET_KEY — Stripe API key (use test key locally)
- JWT_SECRET — 32+ char random string

Dev services (via Docker Compose):
- PostgreSQL: localhost:5432
- Redis: localhost:6379

Start services: `docker compose up -d`
Check health: `/env-doctor`

Example

User clones a repo and runs /env-doctor

Expected output:

🏥 Environment Health Report
════════════════════════════

✅ Node.js 22.16.0 (matches .nvmrc: 22)
✅ npm 10.8.0
✅ node_modules installed, package-lock.json in sync

⚠️  Missing env vars (1):
   - ANTHROPIC_API_KEY
     Not in .env — copy from .env.example and fill in your key
     Get one: https://console.anthropic.com/

❌ PostgreSQL not reachable on localhost:5432
   Fix: docker compose up -d db
   (or set DATABASE_URL to point to your existing Postgres instance)

✅ Redis reachable on localhost:6379
✅ No port conflicts on :3000, :3001

─────────────────────────────
1 blocker | 1 warning
Next step: docker compose up -d db