marketing-ideas
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill consists entirely of informational Markdown content and contains no executable scripts (Python, Node.js, Shell, etc.), which eliminates common technical attack vectors like remote code execution or shell injection.
- [DATA_EXPOSURE]: The instructions direct the agent to check for a project-specific file (
.agents/product-marketing-context.md) to gain context about the user's product. This is a localized file access intended for personalization and does not attempt to access sensitive system files (e.g., SSH keys, AWS credentials) or environment variables. - [INDIRECT_PROMPT_INJECTION]: The skill has a potential surface for indirect injection by reading the contents of the
.agents/product-marketing-context.mdfile. - Ingestion points: The agent reads context from the
.agents/product-marketing-context.mdfile specified inSKILL.md. - Boundary markers: None are specified; the agent is instructed to read and use the context directly.
- Capability inventory: The skill possesses no capabilities for network communication, file writing, or command execution.
- Sanitization: No sanitization is performed on the ingested context.
- Context: Given that the skill lacks any impactful capabilities beyond generating text, the risk associated with this ingestion point is negligible.
Audit Metadata