search-expert
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes content from the public web via the
zai-web-searchtool. - Ingestion points: Untrusted data enters the agent's context through search results in SKILL.md.
- Boundary markers: The skill includes prompt-based guidelines for source verification and cross-referencing to mitigate misinformation or injected instructions.
- Capability inventory: Capabilities are restricted to information retrieval and reporting; no high-risk tools for file system or network manipulation (beyond the search tool) are present.
- Sanitization: The instructions focus on manual verification and synthesis by the agent rather than technical sanitization of the input data.
Audit Metadata