feature-to-benefit
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill consists entirely of markdown instructions (SKILL.md) and JSON test prompts (test-prompts.json). There are no scripts, binaries, or executable commands provided.
- [SAFE]: The instructions focus on marketing and sales communication methods based on "The Copywriter's Handbook" by Robert Bly. No attempts to bypass safety filters, escalate privileges, or establish persistence were detected.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided product descriptions and features. While this represents a data ingestion surface, the skill lacks dangerous capabilities such as file system writing, network requests, or code execution. The risk of malicious instructions in the input data affecting the host system is negligible.
- Ingestion points: User-provided product parameters and descriptions (e.g., in test-prompts.json).
- Boundary markers: The instructions use structured sections (R, I, A1, A2, E, B) to guide the agent, though explicit delimiters for user data are not defined.
- Capability inventory: No subprocess calls, network operations, or file-write capabilities are present.
- Sanitization: Standard LLM text processing is used without specific external sanitization scripts.
Audit Metadata