patentfig

Generate patent-office-compliant line art figures and vectorize drawings via the PatentFig AI API.

30.5k|3.5k|Updated Jul 4, 2025
One-click install
npx skills add https://github.com/davila7/claude-code-templates --skill patentfig
Or copy as Structured Prompt for Agentβ–Ό
Please help me install this Agent Skill.
Skill: patentfig
Source: https://github.com/davila7/claude-code-templates/tree/main/cli-tool/components/skills/creative-design/patentfig
Command: npx skills add https://github.com/davila7/claude-code-templates --skill patentfig

SYSTEM DOCUMENTATION & REQUIREMENTS

πŸ’‘ This Skill includes references (resource) components.

What problem does it solve?

Preparing patent figures that meet USPTO/EPO filing requirements is slow and specialized work. This Skill generates patent line art from text prompts, vectorizes existing drawings into CAD-ready formats, and converts images into filing-ready TIFF/PDF/PNG output through the PatentFig AI REST API.

Core Features & Use Cases

  • Figure Generation: Create patent line art from a text prompt in PNG or stroke-based SVG, with optional numeric reference callouts (100, 102, ...) and up to 4 reference images.
  • Vectorization & Enhancement: Convert raster drawings to SVG, DXF, or vector PDF using lineart or trace engines, and AI-upscale images at 2x or 4x with 300/600 DPI metadata.
  • Filing-Ready Conversion: Export images as flattened TIFF, PDF, or PNG at 300 or 600 DPI for patent office submission.
  • Use Case: An inventor describes an exploded view of a wireless earbud charging case; the Skill generates a labeled SVG figure, then converts it to a 600 DPI TIFF ready for a USPTO filing.

Quick Start

Set the PATENTFIG_API_KEY environment variable, then ask the assistant to generate a labeled SVG patent figure of your invention using the PatentFig API.

Frequently Asked Questions about patentfig

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a patent figure from a text description?β–Ό

Send a POST request to the PatentFig /figures endpoint with a prompt describing the view and parts to draw. Set output to svg for stroke-based line art or png for raster, and labeled to true for numeric reference callouts.

How do I convert a drawing to DXF for CAD?β–Ό

POST the image URL or file to the /vectorize endpoint with format set to dxf and engine set to lineart. The lineart engine redraws the image as clean single-stroke patent line art suitable for CAD import.

What image formats does the PatentFig API accept?β–Ό

The vectorize, enhance, and convert endpoints accept PNG, JPEG, WebP, and TIFF files up to 10 MB. You can pass a public imageUrl in JSON or upload the file directly as multipart/form-data.

Why did my PatentFig API call fail with a 500 error?β–Ό

A 500 GENERATION_FAILED error means the model or processing pipeline failed, often because the input does not suit the endpoint, such as sending a photo for line-art vectorization. Retry once, or use /figures with the photo as a reference image instead.

Does the PatentFig API charge credits for failed requests?β–Ό

No, credits are charged only on successful calls, so retrying a failed request never double-bills. You can check your balance anytime with the free GET /credits endpoint before running batch jobs.