darwin-search
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface that processes external content from search results, which is a common vector for indirect prompt injection.
- Ingestion points: The skill retrieves content through the
search_darwin_capabilitiestool which queries the external Darwin platform (SKILL.md). - Boundary markers: The instructions do not specify the use of delimiters or specific markers to isolate tool outputs from the agent's internal reasoning context.
- Capability inventory: The skill is strictly limited to read-only search operations and documentation retrieval. It explicitly delegates execution actions to a separate skill (
darwin-act), which mitigates the risk of automated exploitation of injected instructions. - Sanitization: There are no specific instructions for sanitizing or escaping the content returned from the Darwin API before it is processed by the agent.
Audit Metadata