referral-program
Pass
Audited by Gen Agent Trust Hub on Mar 7, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface by incorporating untrusted data into its context.
- Ingestion points: The instructions in
SKILL.mddirect the agent to read and utilize data from.agents/product-marketing-context.mdor.claude/product-marketing-context.mdwhen gathering information about the user's product. - Boundary markers: The skill does not define specific delimiters or instructions (e.g., 'ignore embedded commands') to separate the ingested context file content from the agent's core instructions.
- Capability inventory: While the skill itself is purely informational, the agent using it is capable of interacting with external marketing tools (such as Rewardful, Tolt, and Stripe) and other tactical skills (like launch-strategy).
- Sanitization: There is no evidence of validation, filtering, or sanitization applied to the content of the context files before they are processed by the agent.
Audit Metadata