cafci-fondos-comunes-argentina
Pass
Audited by Gen Agent Trust Hub on Feb 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS] (LOW): The skill makes network requests to an external, non-whitelisted domain (anduin.ferminrp.workers.dev) to retrieve fund information.
- [COMMAND_EXECUTION] (SAFE): The skill uses curl and jq to fetch and format data. These commands are used as intended for data retrieval and do not involve executing untrusted scripts or piping to shells.
- [DATA_EXFILTRATION] (LOW): Performs network operations to a non-whitelisted domain. While no sensitive local data is transmitted, the destination is not an approved trusted source.
- [PROMPT_INJECTION] (LOW): The skill is susceptible to Indirect Prompt Injection (Category 8). 1. Ingestion points: Financial data returned from the external API. 2. Boundary markers: Absent. No specific delimiters are defined to separate untrusted data. 3. Capability inventory: Uses curl and jq to process and display information. 4. Sanitization: Absent. The skill does not validate or sanitize the contents of the API response before presentation.
Audit Metadata