Blindwatermark Skill

Simplified Chinese

A publish-ready Codex/OpenCode skill for blind image watermark workflows.

This repository is intentionally small: it is for distributing the skill itself, not for shipping a Python package, demo application, example project, or extra documentation bundle.

What It Contains

• SKILL.md: the assistant-facing skill instructions.
• INSTALL.md: how to install the skill into a local skills directory.
• QUICKSTART.md: the shortest way to start using the skill.

The skill helps an assistant guide watermark embedding, extraction, verification, metadata handling, and troubleshooting. It also reminds the assistant to avoid unsupported claims such as cryptographic proof or guaranteed watermark recovery.

Repository Layout

blindwatermark-skill/
|-- SKILL.md
|-- README.md
|-- README.zh-CN.md
|-- INSTALL.md
|-- QUICKSTART.md
`-- .gitignore

Installation

Copy this directory into your Codex/OpenCode skills directory:

mkdir -p ~/.codex/skills
cp -r blindwatermark-skill ~/.codex/skills/

For project-local use:

mkdir -p .codex/skills
cp -r blindwatermark-skill .codex/skills/

Then ask your assistant to use the blindwatermark-skill skill when working on watermarking tasks.

Usage

After installation, ask your assistant for help with blind watermark tasks, for example:

Use blindwatermark-skill to help me embed a hidden asset ID into this image and verify extraction.

The assistant will provide the runtime commands or scripts needed for the user's actual workspace.