guillaumemeyer/watermarks-remover

What it solves

This project provides tools to remove AI-generated provenance marks and watermarks from text and files. It is designed for users who own their content and wish to strip these markers for privacy or hygiene purposes, targeting marks from vendors like Claude, Gemini (SynthID-Text), and OpenAI, as well as open-LLM schemes.

How it works

The system operates across three primary layers of removal:

• Layer A (Deterministic): Uses Python scripts to strip invisible Unicode characters, exotic spaces, and tag characters.
• Layer B (Statistical): Handles token-sampling watermarks through agent-led rewrites or a specific rewrite hook.
• Layer C (Files): Strips C2PA, EXIF, XMP, and document properties from a wide variety of file formats including images (PNG, JPEG, WebP, etc.), documents (PDF, DOCX, XLSX, etc.), and media (MP4, WAV, MP3, etc.).

It is delivered as a standalone HTTP service (built with Python stdlib) and as "skills" that can be integrated into AI agents like Claude Code, Cursor, and Grok. It can also be used via direct CLI scripts.

Who it’s for

Content owners who want to remove AI provenance markers from their files and text, and developers integrating these capabilities into AI agent workflows.

Highlights

• Multi-Vendor Support: Targets marks from Claude, Gemini/SynthID, OpenAI, and open-LLM schemes.
• Agent Integration: Ready-made skills for Claude Code, Cursor, and Grok, including a deterministic "hook" that can automatically clean files as an agent writes them.
• Broad File Support: Handles a massive range of formats from images and PDFs to Office documents and audio/video files.
• ListNode-based Service: Provides a REST API for inspecting and cleaning files in batch or individually.
• Zero-Dependency Core: The core service and scripts rely only on the Python 3.10+ standard library.