Imported from Durgeshwar-AI/pdfToPng (AGENTS.md). Install upstream with npx skills add Durgeshwar-AI/pdfToPng. Copyright stays with the author.

AGENTS.md

Architecture

Full-stack app: Flask backend (backend/) + React/Vite frontend (frontend/). Not a monorepo — no workspace tooling. Each side has its own package.json / requirements.txt and runs independently.

• Backend entry: backend/main.py → app/__init__.py creates Flask app, registers blueprints
• Frontend entry: frontend/src/main.jsx → App.jsx (React Router v7)
• API base URL: http://localhost:5000 (set via VITE_API_URL env var)

Hard Rules (from CONTRIBUTING.md)

1. No persistent storage — server processes files in-memory only, never writes to disk/DB
2. No external APIs — all processing is local via installed libraries
3. Only file-manipulation features — format conversion, compression, resize, merge, split, etc.

Violations will not be merged.

Dev Commands

Backend

cd backend
python -m venv venv
venv\Scripts\activate      # Windows
pip install -r requirements.txt
python main.py             # Runs on :5000

Frontend

cd frontend
npm install
npm run dev                # Runs on :5173

Docker (recommended for first setup)

docker-compose up --build  # Frontend :5173, Backend :5000

Lint (frontend only)

cd frontend && npm run lint

There are no test suites, typecheck, or formatter commands configured in this repo.

Code Conventions

• Python: 4-space indent (PEP 8). Blueprints in backend/blueprints/, registered in app/__init__.py. Each blueprint = one file = one feature area.
• JS/JSX: 2-space indent. Functional components + hooks only. React 19, Vite 7, Tailwind CSS v4.
• Line endings: LF (enforced by .editorconfig)
• Unused vars: ESLint ignores vars matching ^[A-Z_] (component names, constants)

Backend Patterns

• Blueprints use a shared validate_uploaded_file() + validate_image_file() flow from utils/validators.py
• The @process_image_request decorator in utils/decorators.py handles upload validation, image loading, and cleanup for image endpoints
• All responses use utils.helpers.error() for error responses and send_file_and_cleanup() for file downloads
• Force gc.collect() after processing large files

Frontend Patterns

• Tool pages live in frontend/src/pages/ — one page per feature
• Shared FileUploadArea component and useFileUpload hook for upload flows
• ToolPageTemplate provides consistent layout for tool pages
• Tailwind CSS v4 via @tailwindcss/vite plugin (no tailwind.config.js — uses CSS-first config)
• PDF.js (pdfjs-dist) used client-side for PDF rendering; backend /convertPng exists but UI defaults to client-side

Adding a New Feature

1. Create blueprint in backend/blueprints/, register it in app/__init__.py
2. Create page in frontend/src/pages/, add route in App.jsx
3. Add tool metadata in frontend/src/data/toolsData.jsx
4. Update Readme.md (required for PR acceptance per CONTRIBUTING.md)

Available Skills (.agents/skills/)

Deployment

• Render: render.yaml at root — backend only (gunicorn, 1 worker, 120s timeout)
• Vercel: frontend/vercel.json — frontend static build
• No CI/CD workflows configured (.github/ is empty)