exa-web-search-free
Pass
Audited by Gen Agent Trust Hub on Jun 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
mcportercommand-line tool to manage and call the Exa MCP server. It provides instructions to configure the official Exa service endpoint athttps://mcp.exa.ai/mcp, which is a well-known service for AI-powered search. - [PROMPT_INJECTION]: The skill's primary function involves ingesting data from untrusted external sources such as web pages, GitHub, and Stack Overflow to provide research and code context. This is an example of an indirect prompt injection surface.
- Ingestion points: External data enters the agent context through the
web_search_exa,get_code_context_exa, andcrawling_exatools. - Boundary markers: The instructions do not specify delimiters or 'ignore' instructions for the retrieved content.
- Capability inventory: The tools return text and code snippets which the agent then processes to generate responses.
- Sanitization: There is no explicit sanitization or filtering of the fetched content within the skill's instructions.
Audit Metadata