budget-pacing-monitor
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data (campaign report exports/CSVs) which serves as an ingestion point for untrusted data. The instructions explicitly warn the agent to treat these files as untrusted and to never execute instructions embedded in them (e.g., in campaign names or labels). While the attack surface exists, the risk is mitigated by these specific boundary instructions.
- [COMMAND_EXECUTION]: The skill uses a local script
ledger.pyviapython3to record and trend data. This is a standard internal tool usage for the skill's state management and does not involve arbitrary command execution from user input. - [DATA_EXPOSURE]: The skill writes to local memory paths (e.g.,
memory/ad/budget-pacing-monitor/) andmemory/open-loops.md. These are standard persistence mechanisms for the agent's internal state and do not involve exfiltration to external domains.
Audit Metadata