Deploy

Input

• An app or service in the current repo, plus a target such as Cloudflare, VPS, or a mix of both.
• Use explicit input first; otherwise infer from context, config, docs, selected files, or branch.
• Safest default: define and document the deploy path before automating it.

Workflow

1. Inspect the app. Identify runtime, build command, start command, ports, env vars, storage, database, background jobs, and health checks.
2. Detect deploy path. Read existing deploy docs, CI/CD, Dockerfiles, platform config, infra files, and package scripts.
3. Choose route. Use Cloudflare, VPS, or mixed deployment based on app needs and user preference.
4. Plan setup if missing. Confirm target, domain/DNS, environments, secrets, build, runtime, database/migrations, logs, monitoring, backups, rollback, and CI/CD trigger.
5. Apply best practices. Prefer least-privilege secrets, HTTPS, non-root processes where applicable, firewall or platform access controls, health checks, logs, rollback path, and documented env vars.
6. Configure deployment. Add or update platform config, CI/CD, Docker/systemd/service files, scripts, and docs using repo conventions.
7. Validate. Run local checks, build, deploy dry run where available, smoke test, health check, logs check, and rollback notes.
8. Document. Keep internal developer and operator deployment notes under docs/development/, such as docs/development/deployment.md, docs/development/ci.md, docs/development/operations.md, or docs/development/environment.md.