Detect indirect prompt injection (goal hijack). Instructions hidden in "external" content (documents, RAG, web) that the agent processes. Use when the agent has document/RAG/web/file input.
安装
npx skills add Tencent/AI-Infra-Guard --skill indirect-injection-detectionnpx skills add Tencent/AI-Infra-Guard --skill indirect-injection-detection -a claude-code -gnpx skills add Tencent/AI-Infra-Guard --skill indirect-injection-detection -a codex -ggemini skills install https://github.com/Tencent/AI-Infra-Guard.git --path agent-scan/agent_scan/prompt/skills/indirect-injection-detectiongh skill install Tencent/AI-Infra-Guard indirect-injection-detectiongit clone --depth 1 https://github.com/Tencent/AI-Infra-Guard /tmp/Tencent__AI-Infra-Guard
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/Tencent__AI-Infra-Guard/agent-scan/agent_scan/prompt/skills/indirect-injection-detection ~/.claude/skills/indirect-injection-detection # Claude Code
cp -R /tmp/Tencent__AI-Infra-Guard/agent-scan/agent_scan/prompt/skills/indirect-injection-detection ~/.agents/skills/indirect-injection-detection # Codex, Cursor, Gemini CLI, Copilot, OpenCode