brand-protection
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and analyze content from untrusted external sources, specifically sites suspected of phishing or impersonation, which creates a surface for indirect prompt injection attacks. \n
- Ingestion points: The 'Evidence Collection Checklist' in SKILL.md requires the agent to document URLs, site content, and WHOIS data from potentially malicious target websites. \n
- Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when processing this external data. \n
- Capability inventory: While primarily instructional, the agent's ability to read and process site content for reporting purposes could be exploited if those sites contain malicious instructions designed to manipulate the agent. \n
- Sanitization: The skill lacks procedures for sanitizing or validating the information retrieved from external sources before it is analyzed or included in reports.
Audit Metadata