SkillEngenhariaRevisão

Agentic Actions Auditor

Audit GitHub Actions and CI workflows for AI/agent-related risks - prompt injection vectors, secret exposure, untrusted code execution.

Ações
PerfilDev
ProfundidadeAlta
Idiomaen-US
Objetivo

Em uma frase.

Use this skill when reviewing CI/CD workflows that include AI agents, LLM calls, or other agentic actions. Audits for prompt injection vectors (PR titles, comments, file contents), secret exposure paths, and untrusted-code execution patterns.

Constelação

Onde
ela vive.

Bundles que incluem