designer_skill
- Python
1
GitHub Stars
2
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill digitalocean-labs/do-app-platform-skills --skill designer- README.md2.0 KB
- SKILL.md8.9 KB
Overview
This skill transforms natural language application descriptions or repository contents into production-ready DigitalOcean App Platform specifications. It generates .do/app.yaml, an optional .do/deploy.template.yaml for public repos, and a .env.example to capture required environment variables. Use it to design multi-component architectures, monorepo layouts, and multi-environment setups quickly and consistently.
How this skill works
Provide a natural language brief or point this skill at a repository. It gathers requirements (framework, database, background jobs, routing) and decomposes the app into App Platform components: services, workers, jobs, static_sites, and databases. The skill applies opinionated defaults (instance size, health checks, regions, deploy-on-push), fills env var placeholders and ingress rules, and validates the resulting YAML against App Platform constraints.
When to use it
- Design a new app from a product description or architecture brief
- Create an app.spec from an existing repository (auto-detect language and Dockerfile)
- Add a Deploy to DigitalOcean button for a public GitHub repo
- Set up API + frontend routing and internal service URLs
- Generate environment-portable configs for local, Docker Compose, and App Platform
Best practices
- Decompose apps into clear components: services for HTTP, workers for background, jobs for scheduled tasks
- Bind services to process.env.PORT and avoid hardcoded URLs; use ${service.PRIVATE_URL} and bindable vars
- Prefer Dockerfile builds if present for maximum reproducibility; otherwise infer build from package files
- Start small: apps-s-1vcpu-1gb and single instance for development, use managed dev DBs by default
- Use health endpoints (/health or /healthz) and declare ingress.rules for explicit routing
Example use cases
- Turn a product brief — "React frontend + Node API + Postgres" — into a full .do/app.yaml with ingress and env mappings
- Scan a monorepo and produce separate components for /api, /worker, and /frontend with correct source_dir settings
- Add a deploy.template.yaml and README button so contributors can deploy the app to DigitalOcean with one click
- Generate .env.example with RUN_TIME and BUILD_TIME variable placeholders and bindable database references
FAQ
Yes. It checks for Dockerfile, package.json, requirements.txt/pyproject.toml, go.mod and infers builds and component types, including monorepo layouts.
Does it produce production-ready defaults?
It applies conservative, opinionated defaults (instance size, deploy-on-push, region nyc, dev databases) that you should review and adjust for production scale and compliance.
Can I get a one-click Deploy to DigitalOcean button?
Yes. For public GitHub repos it generates .do/deploy.template.yaml and the README button URL template required for the Deploy to DO flow.