finance-ops
Pass
Audited by Gen Agent Trust Hub on Apr 16, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill defines a process for ingesting and analyzing external data from sources such as Fakturoid CSV exports, Stripe records, and email receipts. This presents a surface for indirect prompt injection where malicious instructions could be embedded within financial data to influence the agent's behavior. The risk is limited as the skill does not request tools with significant capabilities (e.g., network access or shell execution).
- Ingestion points: Data entering the context from Fakturoid (CSV), Stripe, and email receipts (documented in the 'Zdroje dat' section).
- Boundary markers: No explicit delimiters or 'ignore instructions' warnings are included to wrap the ingested external data.
- Capability inventory: The skill lacks capabilities for network operations, shell command execution, or file writing.
- Sanitization: There are no provided steps for validating or escaping external data before it is interpolated into the analysis report template.
Audit Metadata