product-marketing
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill scans project files (such as READMEs, landing pages, and package.json) to generate marketing summaries. This ingestion of external project data creates a potential surface for indirect prompt injection if those files contain malicious instructions meant to influence the agent. The risk is minimized by the skill's workflow, which requires the user to review and correct the draft before it is saved to the filesystem.
- Ingestion points: Codebase files scanned during the 'Auto-draft' workflow, including README, landing pages, marketing copy, about pages, meta descriptions, and package.json.
- Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the ingestion prompts.
- Capability inventory: The skill writes the finalized context to
.agents/product-marketing.md. - Sanitization: No specific content sanitization is mentioned; the skill relies on human-in-the-loop review to ensure accuracy and safety.
- [SAFE]: The skill's primary functionality is restricted to reading project documentation and writing to a dedicated configuration directory. There is no evidence of unauthorized network activity, credential harvesting, or code obfuscation.
Audit Metadata