Show HN: Watermarks Remover: Clean LLM watermarks from text and files (github.com)

A new tool called **Watermarks Remover** has been introduced, designed to strip multi-vendor AI provenance marks from text and various file formats, enhancing privacy and content hygiene. This service operates over HTTP, allowing users to integrate it with their systems without needing to install Python directly. It can remove both invisible markers and statistical watermarks, making it compatible with a wide range of file types, including images (PNG, JPEG, etc.) and documents (PDF, DOCX, etc.). The skill can be easily installed through different platforms, including Claude Code and Cowork, making it accessible for diverse use cases. The significance of this development lies in its potential to enhance user control over their digital content, particularly as concerns about AI watermarking and data privacy grow. This tool allows content creators and users to cleanse their files of these identifiers, which could be crucial for safeguarding intellectual property and maintaining originality. Technically, it leverages deterministic scripts for watermark removal, facilitating seamless integration into user workflows while ensuring that content remains untainted for personal or project-based interactions. This aligns with the broader AI/ML trend towards fostering privacy-conscious tools while navigating the evolving landscape of AI-generated content.