content-strategy
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions involve processing untrusted data from multiple external sources, which could contain hidden instructions intended to influence the agent's behavior.
- Ingestion points: The skill analyzes user-provided keyword exports, sales/customer call transcripts, survey responses, and content from public forums such as Reddit and Quora (specified in
SKILL.mdunder "Content Ideation Sources"). - Boundary markers: Absent. There are no instructions provided to the agent to treat these external inputs as untrusted data or to use delimiters to prevent instruction injection.
- Capability inventory: The skill possesses capabilities for reading local project files (
.agents/product-marketing.md) and performing web searches for research, creating a path where malicious content from a forum could influence those operations. - Sanitization: Absent. The skill lacks guidance on sanitizing or filtering input from transcripts or forum posts before processing.
Audit Metadata