For AI agents

This project is built to be used and improved by AI coding agents as much as by people. Three steps and your agent knows the whole library:

setup-ai writes the library skill where each agent looks for it (.claude/skills/, AGENTS.md, .cursor/rules/, .github/copilot-instructions.md). Running it again updates the same block instead of duplicating it. Without --agent it detects the agents already used in the project.

Using the library from an agent

• Agent skill: the npm package ships skills/mri-fivem-liquid-glass/SKILL.md, a ready skill in the Claude Code / Agent Skills format. Copy it into your agent's skills folder, or point the agent at node_modules/mri-fivem-liquid-glass/skills/mri-fivem-liquid-glass/SKILL.md.
• /llms.txt: a short map of the docs and the API for language models; /llms-full.txt has every guide page in one file.
• JSON API: options, attributes and presets as data, always generated from the source.

Improved it? Send it back

If your agent fixed a bug, tuned a shader or added a feature while using this library in a project, open a pull request upstream instead of keeping a private patch. Everyone gets the improvement and your project keeps getting updates without conflicts.

1. Read AGENTS.md in the repository: how to build, test and what reviewers expect.
2. Keep the change focused and explain what was seen in game versus in the browser.
3. Open the pull request at github.com/mur4i/mri-fivem-liquid-glass.

Changes to docs, examples and the community registry merge automatically when the checks pass. Changes to the library code get a quick review from a maintainer.

Ground rules for agents

• Target the FiveM browser: Chromium 103 and WebGL2. No newer browser APIs.
• Never copy code from closed or paid resources.
• Do not claim it was tested in game unless it was.