referral-program
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWPROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection (LOW): The skill instructions direct the agent to ingest data from an external context file, which could theoretically contain adversarial instructions. 1. Ingestion point: .claude/product-marketing-context.md (referenced in SKILL.md). 2. Boundary markers: Absent; there are no instructions to the agent to delimit or ignore instructions within that file. 3. Capability inventory: No high-risk capabilities such as command execution, network access, or file-writing are present in the provided skill files. 4. Sanitization: Absent. The risk is limited to influencing the agent's textual advice or internal reasoning.
Audit Metadata