The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill's configuration instructions involve downloading and executing the exa-mcp-server package from the NPM registry via npx. This is the standard method for deploying the official search server for the Exa service.
[COMMAND_EXECUTION]: The setup instructions utilize the npx command-line tool to execute the MCP server in the local environment.
[SAFE]: The skill provides instructions for managing authentication via the EXA_API_KEY environment variable using placeholders, which aligns with standard security best practices for secret management.
[SAFE]: Regarding indirect prompt injection (Category 8):
Ingestion points: Untrusted data enters the agent context through the web_search_exa and get_code_context_exa tools, which retrieve content from the public web and code repositories.
Boundary markers: The skill does not explicitly define delimiters or specific "ignore embedded instructions" warnings for the returned search content.
Capability inventory: The skill itself is restricted to read-only search operations and does not include file-system write or arbitrary command execution capabilities.
Sanitization: No explicit sanitization or filtering of the external search results is performed by the skill instructions.

exa-search