research
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources, creating a potential surface for indirect prompt injection.
- Ingestion points: Technical documentation retrieved via
mcp__context7__query-docsand web/code search results frommcp__exa__web_search_exaandmcp__exa__get_code_context_exainSKILL.mdandreferences/tool-usage.md. - Boundary markers: The instructions lack explicit delimiters or warnings to the agent to disregard instructions potentially embedded within the retrieved external data.
- Capability inventory: The agent has access to
context7,exa, andsequential-thinkingMCP tools to perform research and synthesis. - Sanitization: No explicit sanitization, filtering, or validation of external content is defined before the agent processes and synthesizes the information.
Audit Metadata