customer-research
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is designed to ingest and process data from external, untrusted sources, creating a surface for indirect prompt injection.
- Ingestion points: The skill instructions in
SKILL.mdandreferences/source-guides.mdinvolve retrieving and analyzing content from Reddit, G2, LinkedIn, YouTube, and customer support tickets. - Boundary markers: The provided instructions do not include specific delimiters or guidance for the agent to distinguish between its instructions and the potentially untrusted content being analyzed.
- Capability inventory: The skill requires the agent to browse the web and read files to perform its research and synthesis tasks.
- Sanitization: There is no evidence of specific sanitization or filtering logic to handle malicious instructions that might be embedded in the external reviews or social media posts being processed.
Audit Metadata