backend-server-verification

# Backend Server Verification Before running any API tests against a backend, verify the server is actually running and reachable. ## Before Starting 1. **Check if target port is in use**: `lsof -i :PORT` (or equivalent). 2. **If in use**: Kill the process or document using an alternate port. Do not assume "server stuck" or blame Redis/OpenAI—port conflict is a common cause of failures. ## Start Command - **Always use an explicit directory** so CWD is unambiguous across tool invocations. - Example: `cd /absolute/path/to/backend && npm run dev` (prefer absolute path). ## After Start 1. **Wait briefly** (e.g. a few seconds for the process to bind). 2. **Confirm readiness**: `curl -f http://localhost:PORT/api/health --max-time 5`. 3. **If curl fails**: Read the terminal output for "port in use" or bind errors before retrying with long timeouts. ## Two-Step Pattern (Recommended) - **Invocation 1**: Start the server (background or in a separate terminal). - **Invocation 2**: Run curl with `--max-time 5` to verify; do not run long curls until health succeeds. ## Integration - Use with `dev-server-lifecycle-management` and `port-conflict-resolution` for port detection and conflict handling. - Use with `pre-flight-checklist` / `pre-implementation-check` for backend tasks.