Runs NVIDIA garak probe suites (jailbreak, prompt injection, data leakage, toxicity, and more) against an LLM endpoint - Hugging Face models, OpenAI-compatible APIs, or Bedrock - then interprets the resulting hit-rate report for triage.
安装
npx skills add mukul975/Anthropic-Cybersecurity-Skills --skill red-teaming-llms-with-garaknpx skills add mukul975/Anthropic-Cybersecurity-Skills --skill red-teaming-llms-with-garak -a claude-code -gnpx skills add mukul975/Anthropic-Cybersecurity-Skills --skill red-teaming-llms-with-garak -a codex -ggemini skills install https://github.com/mukul975/Anthropic-Cybersecurity-Skills.git --path skills/red-teaming-llms-with-garakgh skill install mukul975/Anthropic-Cybersecurity-Skills red-teaming-llms-with-garakgit clone --depth 1 https://github.com/mukul975/Anthropic-Cybersecurity-Skills /tmp/mukul975__Anthropic-Cybersecurity-Skills
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/mukul975__Anthropic-Cybersecurity-Skills/skills/red-teaming-llms-with-garak ~/.claude/skills/red-teaming-llms-with-garak # Claude Code
cp -R /tmp/mukul975__Anthropic-Cybersecurity-Skills/skills/red-teaming-llms-with-garak ~/.agents/skills/red-teaming-llms-with-garak # Codex, Cursor, Gemini CLI, Copilot, OpenCode