product-marketing
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill includes a feature to automatically draft marketing context by scanning the local codebase, which involves processing untrusted data.
- Ingestion points: The skill reads files including
README.md,package.json, landing page source code, and existing documentation to extract product and audience information (Step 2 inSKILL.md). - Boundary markers: The instructions do not define clear delimiters or use "ignore embedded instructions" warnings when the agent interpolates the content from the codebase into its drafting process.
- Capability inventory: The skill possesses the capability to write to the filesystem at
.agents/product-marketing.mdand reads files across the project directory. - Sanitization: There are no explicit sanitization or filtering steps mentioned for the extracted content; the skill relies on a human-in-the-loop verification step where the user reviews the draft before it is saved.
Audit Metadata