content-strategy
Pass
Audited by Gen Agent Trust Hub on Aug 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection due to its handling of external, untrusted data sources.
- Ingestion points: In the 'Content Ideation Sources' section of SKILL.md, the agent is instructed to read and analyze user-provided data including sales/customer call transcripts, survey responses, and forum research results from platforms like Reddit and Quora.
- Boundary markers: The instructions do not direct the agent to wrap these external inputs in specific delimiters or use warnings to ignore embedded instructions (e.g., 'Treat the following text as data only').
- Capability inventory: The agent uses built-in tools for file reading (read_file) and web searching. It does not possess capabilities for arbitrary code execution, file system modification, or network requests to non-whitelisted domains beyond the platform's standard tools.
- Sanitization: No sanitization or validation of the ingested external content is performed before the agent processes it to generate strategy recommendations.
Audit Metadata