marketing-referral
Pass
Audited by Gen Agent Trust Hub on Apr 27, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to install the
agent-browserutility from the Vercel Labs GitHub repository. This is a well-known and trusted organization, and the download is part of the intended setup for research capabilities.\n- [PROMPT_INJECTION]: The skill possesses a 'Live Website URL Audit' mode inreferences/shared-patterns.md, creating a surface for indirect prompt injection.\n - Ingestion points: Data enters the agent context from untrusted external URLs during site audits.\n
- Boundary markers: Absent; no delimiters or instructions are used to distinguish external data from system instructions.\n
- Capability inventory: The skill can write marketing strategy files to the local file system and use browser-based tools.\n
- Sanitization: Absent; the skill does not specify any validation or filtering of ingested web content.
Audit Metadata