wachi-validacion-interna

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project records (referred to as "fichas") which likely contain untrusted data from external stakeholders. If these records contain malicious instructions, they could influence the agent's review logic or its use of project management tools.
  • Ingestion points: The primary ingestion point is the project record or contract processed during the review panel described in Fase 1 of SKILL.md.
  • Boundary markers: The instructions do not define explicit data delimiters or "ignore instructions" warnings for the ingested data. The workflow requires presenting each decision to the user, which serves as a manual oversight mechanism.
  • Capability inventory: The skill utilizes tools to interact with workflow systems, including crear_ticket, vincular_adr_ficha, resolver_adr_ficha, actualizar_dor_ficha, transition_ficha, crear_release, and agregar_ficha_a_release.
  • Sanitization: No sanitization or validation logic for external content is mentioned in the instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 04:39 PM
Security Audit — agent-trust-hub — wachi-validacion-interna