ai-search
Pass
Audited by Gen Agent Trust Hub on Nov 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from the web using search and scraping tools, which presents a surface for indirect prompt injection. * Ingestion points: Phase 2 and Phase 4 fetch data from search results and external URLs via Firecrawl and Exa. * Boundary markers: No delimiters or boundary instructions are defined for the ingested data. * Capability inventory: The skill writes audit reports and notes to the local filesystem and executes a context script. * Sanitization: No sanitization of the external web data is specified.
- [COMMAND_EXECUTION]: The skill executes a local Python script (brand_context.py) for context initialization using the system's python3 interpreter.
Audit Metadata