client-management
Pass
Audited by Gen Agent Trust Hub on Mar 4, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The capability
bulk_import_clientstakes acsv_file_urlparameter, allowing the agent to fetch and process data from arbitrary external sources. - [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection via the bulk import mechanism. Ingestion points: Data from external CSV files is ingested into the agent context through the
csv_file_urlparameter. Boundary markers: No delimiters or safety instructions are defined to separate the imported data from the agent's instructions. Capability inventory: The skill possesses capabilities to write to and modify the client database (create/update) based on the ingested data. Sanitization: There is no evidence of sanitization or validation of the content within the downloaded CSV file prior to processing.
Audit Metadata