The Agent Skills Directory

[SAFE]: The skill provides legitimate neural search functionality via the Exa platform, intended for web and technical research tasks.
[EXTERNAL_DOWNLOADS]: Instructions involve using npx to fetch and run the exa-mcp-server package. This is the standard distribution method for this well-known service's MCP implementation and is documented neutrally.
[PROMPT_INJECTION]: The skill ingests untrusted data from the web via web_search_exa and get_code_context_exa (SKILL.md), creating an indirect prompt injection surface. This is an inherent property of search-based skills and is managed by the agent's internal safety guardrails rather than specific skill-level mitigations.

exa-search