dpma-search

Search the German Patent and Trademark Office (DPMA) register programmatically. Retrieves trademark data including status, Nice classifications, and owner information.

Uses Kernel Cloud for headless browser sessions and Playwright for page automation.

How it works

The DPMA register doesn't offer a public API. This tool launches a cloud browser instance, navigates the register, and scrapes structured data from search results and detail pages.

Features

• Full-text trademark search against the DPMA register
• Scrapes detail pages for each result (filing date, status, owner, address)
• Extracts Nice classifications with descriptions
• Optional JSON output for further processing
• Next.js web interface for browser-based searches

Requirements

• Bun runtime
• Kernel Cloud account (for cloud browser sessions)

Installation

git clone https://github.com/Synistic/dpma-search.git
cd dpma-search
bun install

Usage

CLI

bun run search "your trademark query"

# With JSON output
bun run search "your trademark query" --json

Web Interface

bun run dev    # Start Next.js dev server

Stack

• Runtime: Bun
• Frontend: Next.js, Tailwind CSS, Radix UI
• Scraping: Playwright via Kernel Cloud CDP
• Language: TypeScript

License

MIT