openfactura-errores
Pass
Audited by Gen Agent Trust Hub on Nov 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data in the form of API error responses from the Open Factura platform.
- Ingestion points: JSON error objects containing code, message, and details fields (SKILL.md).
- Boundary markers: Present. The agent is instructed to 'Decida por code, nunca por el texto del mensaje' to avoid manipulation by variable message content.
- Capability inventory: API actions such as dte_emitir, dte_obtener, and dte_invalidar (SKILL.md).
- Sanitization: Present. The skill forces logic to rely on enumerated status codes rather than user-controlled or API-controlled message strings.
Audit Metadata