dariocabas/starter-kit
Overview
This skill analyzes a codebase and explains how to start the project locally. It detects the technology stack, package manager, frameworks, required services, environment variables, and available setup scripts. The output is a concise, actionable starter guide with quick dev commands and full setup steps.
How this skill works
The skill scans repository files that indicate language and tools (package.json, requirements.txt, go.mod, Cargo.toml, etc.). It reads lockfiles to infer the package manager, parses dependencies to recognize frameworks, inspects docker-compose for services, and extracts variable names from .env.example. It also looks for setup, migration, seed, or mock scripts to recommend test data and run sequences.
When to use it
- When you ask "how do I initialize the project" or "how to setup"
- When onboarding a new repo and you need quick dev and full setup steps
- When preparing a local environment that may require databases or caches
- When you need a minimal fast-start vs complete reproducible setup
- When documenting a README starter section or developer guide
Best practices
- Never read or expose sensitive files (.env, *secret*, *.key, *.pem) — only use .env.example
- Provide both a minimal dev path (no external services) and a full path (with Docker/services)
- List only environment variable NAMES and group them by category (db, auth, external APIs)
- Prefer explicit commands for package manager, migrations, seeds, and Docker
- Note common framework-specific commands (e.g., start scripts, dev servers, migration tools)
Example use cases
- Generate a one-page starter section for a project README with quick and full commands
- Onboard new developers by listing exact commands to run the app locally and test data seeds
- Audit a repository to identify which services (Postgres, Redis, Mongo) need to run
- Convert repository contents into a reproducible dev environment checklist
- Help support teams reproduce customer issues by outlining setup and seed steps
FAQ
No. It will never read .env, files containing "secret", credentials, or key files. It only reads .env.example and non-sensitive config files.
How does it detect the package manager?
It checks for lockfiles (yarn.lock, package-lock.json, pnpm-lock.yaml, poetry.lock, etc.) and chooses the matching manager.
Will it run code or containers automatically?
No. It only inspects files and produces commands and recommendations. Executing commands remains a manual step.