Summary This project addresses the challenge of converting academic research papers into formal Chinese invention patent applications. It provides AI-powered prompts and structured templates to assist researchers and patent engineers in navigating the complex process of patent drafting, thereby helping to protect valuable intellectual property derived from scientific advancements.

How It Works The system leverages large language models (LLMs) through meticulously crafted "Flash/Pro Prompts" and "Agent Skills." It deconstructs research papers to identify core technical contributions, then reconstructs this information into legally compliant Chinese patent application text. The approach emphasizes strict adherence to the Chinese Patent Law, hierarchical claim structuring, and faithful representation of the original paper's technical content, aiming for a balance between legal protection and original fidelity.

Quick Start & Requirements Basic usage involves copying provided prompts directly into compatible AI chat interfaces (e.g., Codex, Claude Code, Cursor, Windsurf). For advanced features like patent drawing generation, specific LLMs like Gemini 3.1 Pro are recommended. Installation typically involves copying the project's skills/paper2patent/ directory into the AI tool's skill directory. No specific hardware dependencies like GPUs or CUDA versions are listed, but access to capable AI models is essential.

Highlighted Details Compliance: Strictly adheres to the 2020 revision of Chinese Patent Law and its implementation rules. Usability: Designed for "out-of-the-box" use with prompts that eliminate the need to memorize complex patent drafting procedures. Fidelity: Guarantees that all technical content is derived solely from the original paper, without additions, modifications, or distortions. Claims Strategy: Employs a hierarchical claims layout, progressing from broad independent claims to detailed dependent claims, ensuring layered protection. AI-Assisted Drawing: Includes a dedicated prompt for generating patent-compliant drawings using models like Gemini.

Maintenance & Community The project is maintained by "[7toCR]". No specific details regarding community channels (Discord, Slack), active contributors beyond the primary maintainer, sponsorships, or a public roadmap are provided in the README.

Licensing & Compatibility The project is released under the MIT License, permitting free public use. However, it prohibits direct copying to other repositories and claiming authorship. While generally permissive for commercial use and integration into closed-source projects, users should review the specific terms and attribution requirements.

Limitations & Caveats The tool is specifically tailored for generating Chinese invention patents. Its effectiveness is contingent on the quality and patentability of the source research paper, as it strictly avoids adding external information. The patent drawing generation feature relies on specific AI models and configurations, which may have their own limitations or accessibility issues.