customer-research
Pass
Audited by Gen Agent Trust Hub on Jun 14, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection attacks by directing the agent to ingest and synthesize untrusted data.
- Ingestion points: The skill actively directs the agent to gather intelligence from external, attacker-controllable platforms including Reddit, G2, Capterra, YouTube comments, and public forums (SKILL.md, references/source-guides.md).
- Boundary markers: The instructions lack requirements for the agent to use delimiters, data-wrapping, or 'ignore embedded instructions' warnings when processing these external sources.
- Capability inventory: The skill facilitates file system reads (e.g., product-marketing context files) and relies on the agent's web-browsing or search tools to extract research content.
- Sanitization: There are no prescribed sanitization or validation protocols for the content retrieved from online sources before it is analyzed and used to generate deliverables.
Audit Metadata