raygent-skills

Raygent Skills

Ray's Skills for AI Agents, starting with Claude Skills.

Skills

OpenRouter

Use OpenRouter's unified API to call different LLM models with cost and performance tracking.

Features:

• Call any model available on OpenRouter (Claude, GPT, Gemini, Llama, etc.)
• Track response time and cost per request
• Compare models for quality, speed, and cost
• Use modifiers like :nitro (fastest) and :online (web search)

Installation: See Installation below.

Installation

Via Claude Code Plugin System

You can register this repository as a Claude Code Plugin marketplace by running the following command in Claude Code:

/plugin marketplace add rawwerks/raygent-skills

Then, to install the skills:

1. Select Browse and install plugins
2. Select raygent-skills
3. Select raygent-skills (the plugin containing OpenRouter)
4. Select Install now

Manual Installation

1. Download openrouter.zip from the releases or directly from the repository
2. Unzip to your Claude skills directory:

   unzip openrouter.zip -d ~/.claude/skills/
   

Development

Requirements

• Python 3
• skill-creator plugin installed
• jq (for bash scripts)

First-Time Setup

After cloning this repository, run the setup script to install git hooks:

bash scripts/setup_hooks.sh

This installs a pre-commit hook that automatically packages skills when committing.

Auto-Packaging

This repository uses a pre-commit hook to automatically package skills when committing. This ensures .zip files are always up-to-date with source changes.

How it works:

1. When you run git commit, the hook automatically runs
2. All skills (directories with SKILL.md) are packaged
3. Updated .zip files are staged and included in the commit
4. If packaging fails, the commit is aborted

Manual packaging:

# Package all skills
bash scripts/package_all_skills.sh

# Package a specific skill
python3 ~/.claude/plugins/marketplaces/anthropic-agent-skills/skill-creator/scripts/package_skill.py ./openrouter .

Adding New Skills

1. Create a new skill directory with SKILL.md
2. Update .claude-plugin/marketplace.json to include the new skill
3. Commit your changes - the skill will be auto-packaged

Security

Important: Never commit API keys or secrets!

• .env files are gitignored
• API keys should be set as environment variables
• Check .gitignore before committing sensitive files

Contributing

Contributions welcome! Please ensure:

1. Skills follow the Anthropic skill guidelines
2. Skills pass validation (package_skill.py validates automatically)
3. Documentation is clear and includes examples

License

MIT

Author

Raymond Weitekamp (@raygent)