product-marketing
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes user-provided external data, including product descriptions and existing messaging documents. This creates a surface for indirect prompt injection where instructions embedded in those inputs could potentially influence the agent's behavior. Additionally, the instruction to provide 'No warnings, notes, or disclaimers' can be seen as a directive to suppress typical AI safety or meta-informational guardrails in the final output.
- [PROMPT_INJECTION]: Mandatory Evidence Chain for Indirect Prompt Injection:
- Ingestion points: SKILL.md specifies that users should provide existing positioning, messaging, ICP documents, and product descriptions.
- Boundary markers: The instructions do not define clear delimiters or markers to separate user-provided content from the skill's instructions.
- Capability inventory: No scripts or tools are utilized; the skill only generates text output.
- Sanitization: No sanitization or filtering logic is specified for the ingested content.
- [SAFE]: The skill includes a positive security instruction to exclude hidden Unicode characters and replace em dashes, which helps prevent character-based obfuscation techniques in the generated output.
- [SAFE]: No scripts, dependencies, or network operations are present in the skill, adhering to a no-code structure that avoids remote code execution or data exfiltration risks.
Audit Metadata