Claude-Code-Project-Structure-Learning-Repository: Command for Claude Code

/deploy

Deploy the Blog API to the target environment.

Usage

/deploy [environment]

• environment — one of staging or production (default: staging)

What this command does

When the user runs /deploy or /deploy staging or /deploy production, execute the following steps in order. Stop and report if any step fails.

---

Step 1 — Pre-flight checks

Run these checks before touching anything. If any fail, abort and tell the user what failed:

# 1a. Make sure the working tree is clean
git status --porcelain
# Expected: no output. If there is output, say "Uncommitted changes found — commit or stash first."

# 1b. Make sure we're on the right branch
git branch --show-current
# Expected: "main" for production, any branch for staging.

# 1c. Run the full test suite
pytest arc/tests/ -v --tb=short
# Expected: 0 failures. Do not deploy if tests fail.

# 1d. Run the linter
ruff check arc/
# Expected: exit code 0.

---

Step 2 — Build

# Build a deployable package (wheel)
pip install build
python -m build --wheel

The wheel will appear in dist/. Note the filename for Step 3.

---

Step 3 — Deploy

For staging:

# Push the wheel to the staging server via scp, then restart the service
scp dist/*.whl [email protected]:/opt/blog-api/
ssh [email protected] "
  cd /opt/blog-api &&
  pip install --force-reinstall *.whl &&
  systemctl restart blog-api
"

For production:

# Same as staging but target production server
# IMPORTANT: always confirm with the user before running this block
scp dist/*.whl [email protected]:/opt/blog-api/
ssh [email protected] "
  cd /opt/blog-api &&
  pip install --force-reinstall *.whl &&
  systemctl restart blog-api
"

Always ask the user to confirm before deploying to production. Say: "Ready to deploy vX.Y.Z to PRODUCTION. This will affect live users. Confirm? (yes/no)"

---

Step 4 — Post-deploy smoke test