campaign-analytics
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes marketing campaign and customer journey data from external JSON files, creating a potential surface for indirect prompt injection. \n- Ingestion points: Untrusted data enters the agent context through the input JSON files processed by the attribution_analyzer.py, funnel_analyzer.py, and campaign_roi_calculator.py scripts. \n- Boundary markers: The instructions do not provide delimiters or specific warnings for the agent to ignore instructions that might be hidden within the data fields (e.g., channel names, campaign names, or referrer strings). \n- Capability inventory: The included Python scripts perform mathematical analysis and print text/JSON results; they do not possess capabilities for network communication, file system modification, or dynamic code execution. \n- Sanitization: The analysis scripts do not escape or sanitize string values from the input JSON before including them in output tables and reports, which could allow malicious content to be interpreted as instructions if processed by a downstream agent.
Audit Metadata