菜单
Use when orchestrating headless tooling via MCP in this container (filesystem, fetch, playwright, semgrep, trivy, promptfoo, sqlite, kubernetes). Ensures outputs are saved under /pentest/ and avoids brittle npx cold-starts.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when performing external pentesting (recon, scanning, vuln triage) in this container. Produces reproducible commands and writes all outputs under /pentest/ (recon/, scans/, exploits/, reports/).
Use when testing LLM applications or agentic systems for prompt injection, jailbreaks, tool abuse, data leakage, and eval regressions. Uses garak/promptfoo and the configured MCP servers; writes all artifacts under /pentest/.
| name | mcp-orchestration |
| description | Use when orchestrating headless tooling via MCP in this container (filesystem, fetch, playwright, semgrep, trivy, promptfoo, sqlite, kubernetes). Ensures outputs are saved under /pentest/ and avoids brittle npx cold-starts. |
| license | MIT |
| metadata | {"owner":"mayflowergmbh/kali-ai-redteam"} |
filesystem: read/write under /pentest, /tmp, /var/logfetch: HTTP client (mcp-server-fetch)playwright: headless browser automationmemory: scratch memory for the sessionpromptfoo: run evals/redteamtrivy: run trivy mcp for scanssemgrep: run semgrep-mcp for repo scanssqlite: local db at /pentest/mcp.sqlitekubernetes: cluster inspection (uses kubectl config)/pentest/ (never to $HOME)./pentest/reports/.