Build automated multi-turn adversarial attacks against conversational LLM targets using Microsoft PyRIT's RedTeamingOrchestrator, CrescendoOrchestrator (gradual escalation), and TreeOfAttacksWithPruningOrchestrator (adaptive branching), with scorer feedback loops and persisted conversation memory.
安装
npx skills add mukul975/Anthropic-Cybersecurity-Skills --skill orchestrating-llm-attacks-with-pyritnpx skills add mukul975/Anthropic-Cybersecurity-Skills --skill orchestrating-llm-attacks-with-pyrit -a claude-code -gnpx skills add mukul975/Anthropic-Cybersecurity-Skills --skill orchestrating-llm-attacks-with-pyrit -a codex -ggemini skills install https://github.com/mukul975/Anthropic-Cybersecurity-Skills.git --path skills/orchestrating-llm-attacks-with-pyritgh skill install mukul975/Anthropic-Cybersecurity-Skills orchestrating-llm-attacks-with-pyritgit 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/orchestrating-llm-attacks-with-pyrit ~/.claude/skills/orchestrating-llm-attacks-with-pyrit # Claude Code
cp -R /tmp/mukul975__Anthropic-Cybersecurity-Skills/skills/orchestrating-llm-attacks-with-pyrit ~/.agents/skills/orchestrating-llm-attacks-with-pyrit # Codex, Cursor, Gemini CLI, Copilot, OpenCode