marketing-content-strategy
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which could contain hidden instructions.
- Ingestion points: The instructions in
SKILL.mddirect the agent to analyze multiple types of external data, including keyword exports (from Ahrefs, SEMrush, or Google Search Console), sales and customer call transcripts, survey responses, and search results from public forums like Reddit and Quora. - Boundary markers: The instructions do not define clear delimiters or provide specific guidelines for the agent to distinguish between data and commands when analyzing these external sources.
- Capability inventory: The skill leverages the agent's file-reading and web-search capabilities to ingest this data.
- Sanitization: No explicit instructions are provided for sanitizing, validating, or filtering the external content before it is processed by the agent.
Audit Metadata