Security design review for AI agents, MCP servers, and the cloud they can reach. Use when adding or editing an MCP server, agent config, subagent, system prompt, or tool definition, or when the user asks whether an agent setup is safe or has prompt-injection, tool-poisoning, excessive-agency, or lethal-trifecta risk. Runs attestral scan and explains the findings.
$npx -y skills add attestral-labs/attestral --skill attestral-reviewInstalls into the current project.
Run `npx skills use "https://github.com/attestral-labs/attestral" --skill "attestral-labs/attestral/attestral-review"` and follow the generated skill instructions now. Read its complete output, redirecting it to a temporary file first if necessary. Resolve relative paths from the supporting-files directory it provides.
Use the skills in "https://github.com/attestral-labs/attestral" that are relevant to the current task. Run `npx skills add "https://github.com/attestral-labs/attestral"` and select the relevant skills, then follow their instructions.