Security audit
Security checks for vulnerabilities and agentic risk
This Spanish audit skill is transparent about optionally running user-selected verification commands and includes dry-run and secret-filter safeguards, with no hidden persistence or exfiltration found.
Install only if you want a Spanish-language audit workflow that may run commands from repositories you choose. Keep the default dry-run behavior until you have reviewed the target command, avoid --full-env unless necessary, and run third-party gates in a disposable worktree, container, or VM.
Referenced artifact was not completely inspected
1. **Consentimiento explicito por corrida.** Sin `--yes`, `auditar_repo.mjs` NO ejecuta nada: imprime el plan (comando, destino, estado del env) y sale 0. `veri
Referenced artifact was not completely inspected
NO ejecuta nada: imprime el plan (comando, destino, estado del env) y sale 0. `verificar_declaraciones.mjs` corre siempre sus detectores estaticos (100% lectur
The file's comments, usage text, prompts, and status messages are all written in Spanish, including operational output shown to users. There is no indication that the skill is intentionally region-specific or that users can opt into another language, which creates a language-policy concern under the locale-choice rule.
The closing comment states 'esto no mira si el codigo hace lo que dice que hace' and only checks whether declarations hold up. However, the implementation in detectorVeredicto actually runs a supplied gate command via spawnSync and judges whether its reported verdict matches its exit status, which is a direct behavioral check of what the gate claims about itself. This is an active contradiction in the file's own documentation, not merely an omission.
This skill's natural-language interface is entirely in Spanish, including usage guidance, safety warnings, and verdict messages. The file does not offer any user opt-in or alternative locale, which can violate a language/locale policy requiring user choice.
This code file contains user-facing natural-language comments and assertion labels entirely in Spanish, including terms like "Uso", "EJECUCION OMITIDA", and descriptive test names. Under the policy, forcing a specific language without opt-in or documented locale justification is a natural-language policy concern.
Detected: suspicious.dangerous_exec