gestor-autonomos
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from user-provided CSV files (invoices, Stripe/Substack exports) through
scripts/procesar_facturas.py,scripts/generar_libro.py, andscripts/procesar_stripe.py. - Ingestion points: Data enters the agent's context when reading CSV fields such as 'concepto', 'nombre', or 'email'.
- Boundary markers: No explicit instructions or delimiters are used to warn the agent to ignore instructions potentially embedded within the CSV data fields.
- Capability inventory: The skill uses Python scripts to perform calculations and summarizes the results for the agent to present to the user.
- Sanitization: The scripts do not sanitize or filter textual fields before they are displayed in the agent's output context, creating a potential surface for indirect prompt injection if a malicious invoice or transaction record is processed.
Audit Metadata