deploy.md

name	deploy
description	Run tests, build, and deploy to the specified environment (staging or production). Includes safety checks.
argument-hint	<staging|production>
allowed-tools	Bash, Read

Deploy to: $ARGUMENTS

Follow these steps carefully. Stop and report if any step fails.

Safety Checks (always run first)

1. Confirm the target environment is valid. Accepted values: staging, production. If the value is anything else, stop with an error.

2. Run git status to confirm the working tree is clean (no uncommitted changes). If there are uncommitted changes, stop and ask the user to commit or stash them first.

3. Run git log origin/main..HEAD --oneline. If on a non-main branch deploying to production, warn the user and ask for explicit confirmation before proceeding.

4. For production deployments, confirm with the user by printing: "You are about to deploy to PRODUCTION. Type 'yes' to continue." Wait for the user's response before proceeding.

Pre-deploy Steps

5. Run the full test suite. If any tests fail, stop. Do not deploy broken code.

6. Run the build command for this project:

   • npm run build / yarn build / pnpm build for Node projects.
   • python -m build or equivalent for Python packages.
   • make build if a Makefile target exists.
   • If no build step is detected, note this and skip.

Deploy

7. Run the deploy command for the target environment. Look for deploy scripts in these locations (in order of preference):
   • package.json scripts named deploy:$ARGUMENTS or deploy.
   • Makefile targets named deploy-$ARGUMENTS or deploy.
   • Shell scripts in scripts/ or bin/ matching deploy*.sh.
   • A Dockerfile / docker-compose.yml with a deploy workflow. If no deploy script is found, report what was checked and ask the user to specify the deploy command.

Post-deploy

8. Run a smoke check if one exists (e.g., a health check URL or make smoke-test).

9. Print a summary: environment, branch, commit SHA, timestamp, and whether smoke checks passed.