The Agent Skills Directory

[EXTERNAL_DOWNLOADS] (SAFE): The skill references official SDKs (@vapi-ai/server-sdk, @vapi-ai/web) and common libraries (requests). These are standard dependencies for integrating with the Vapi API.\n- [COMMAND_EXECUTION] (SAFE): Includes example cURL commands for API interaction and a command to add an MCP server (npx -y mcp-remote https://docs.vapi.ai/_mcp/server). These are configuration and usage examples related to the skill's primary purpose.\n- [DATA_EXFILTRATION] (SAFE): Network operations are restricted to the official API endpoint (api.vapi.ai). Sensitive data like API keys are handled through environment variables (VAPI_API_KEY) rather than hardcoding, following security best practices.\n- [CREDENTIALS_UNSAFE] (SAFE): No hardcoded credentials or secrets were found. The skill correctly uses placeholders and environment variables for authentication.\n- [REMOTE_CODE_EXECUTION] (SAFE): The MCP server setup command uses a remote URL from the official service domain (docs.vapi.ai) to provide documentation capabilities to the agent. This is a standard and expected feature of the ecosystem.

create-call