| name | web-stack-planner |
| description | Collect project brief and stack choices, then output canonical project-config.json for downstream web-builder skills. |
| compatibility | codex, claude-code, claude-ai, agent-skills |
| license | Apache-2.0 |
| allowed-tools | ["read_file","write_file","run_terminal_cmd","web_search"] |
| metadata | {"category":"web-builder","stage":"1-planning","pipeline":"web-builder"} |
Web Stack Planner
Objective
Capture a complete web project brief and produce a validated project-config.json
that all downstream web-builder skills consume.
Required Workflow
- Ask the user for project name, project type, target audience, authentication needs, and whether any codebase already exists.
- Present frontend framework choices:
- React
- Next.js
- Vue
- Nuxt
- SvelteKit
- Angular
- Astro
- plain HTML/CSS/JS
- Present backend/API choices:
- Node.js/Express
- FastAPI
- Django
- Laravel
- Go/Gin
- Rails
- Supabase Edge Functions
- serverless-only
- Present database choices:
- PostgreSQL
- MySQL
- MongoDB
- SQLite
- Supabase
- PlanetScale
- Turso
- Redis (as cache)
- Firebase
- Present CSS/design system choices:
- Tailwind CSS
- Shadcn/ui
- Chakra UI
- Material UI
- Ant Design
- Mantine
- DaisyUI
- vanilla CSS
- Ask if they need integrations for auth, payments, email, file storage, and i18n:
- Auth: Clerk, Auth.js, Supabase Auth, custom JWT
- Payments: Stripe, Paddle, LemonSqueezy
- Email: Resend, SendGrid
- Storage: S3, Supabase Storage, Cloudflare R2
- Ask deployment target:
- Vercel
- Netlify
- Fly.io
- Railway
- Render
- AWS
- GCP
- Azure
- DigitalOcean
- Cloudflare Pages
- Run
scripts/stack_planner.py to validate and write the canonical project-config.json.
Execution
python skills/web-stack-planner/scripts/stack_planner.py --input <brief.json> --output <out.json> --format json
Output
project-config.json
- Script result file (
json, md, or csv) with status, summary, artifacts, and details
References