paid-ads
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to ingest and process data from external files, creating a surface for indirect prompt injection attacks where malicious instructions could be embedded in the data.
- Ingestion points: The agent is instructed to read
.claude/product-marketing-context.md(SKILL.md) and theroas_calculator.pyscript is designed to process external JSON campaign files provided via the--fileargument. - Boundary markers: Absent. The instructions do not define delimiters or specific safety warnings to the agent regarding instructions that might be contained within these ingested files.
- Capability inventory: The skill uses external advertising platform integrations (Google Ads MCP tool) and executes a local Python script (
scripts/roas_calculator.py) to perform calculations. - Sanitization: Absent. There is no explicit validation, filtering, or sanitization of the content from the external files before it is processed by the agent or the script.
Audit Metadata