marketing-brain
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from a wide variety of external sources, creating a surface for indirect prompt injection.
- Ingestion points: Data is ingested from HubSpot, Pipedrive, Upsales, Shopify, WooCommerce, Stripe APIs, and external websites via the WebFetch tool.
- Boundary markers: The instructions do not mandate the use of delimiters or safety instructions to separate retrieved external content from the agent's primary instructions.
- Capability inventory: The skill utilizes file-writing capabilities (Write, Edit) and local command execution (Bash).
- Sanitization: The instructions do not specify any sanitization, filtering, or escaping for data retrieved from external integrations before it is summarized or written to documentation nodes.
- [DYNAMIC_EXECUTION]: The skill uses local command execution to process data retrieved from external sources.
- Evidence: Instructions direct the agent to "aggregate with jq" via the Bash tool to process JSON data from Stripe. This involves the dynamic generation of shell commands to handle data from external APIs, which represents a low-severity dynamic execution surface.
Audit Metadata