SAFE Integrations
Connect Google Analytics to your AI tools via MCP. Pick yours below — setup takes under two minutes.
AI Chat Assistants
Claude
Anthropic's AI assistant. Add as a custom connector via Settings → Connectors.
Set up →ChatGPT
OpenAI's AI assistant. Connect via custom MCP connector on Plus, Team, or Enterprise.
Set up →Perplexity
AI search assistant. Add as a custom remote connector with Streamable HTTP transport.
Set up →Mistral Le Chat
European AI assistant. Add as a custom MCP connector via Intelligence → Connectors.
Set up →Langdock
Enterprise AI workspace. Supports Streamable HTTP transport natively.
Set up →Sana
Enterprise employee assistant. Admin adds once; users connect their own Google account.
Set up →LibreChat
Self-hosted multi-user chat. Configure via librechat.yaml with streamable-http type.
Set up →Agent Platforms
CrewAI
Enterprise agent workspace. Requires Streamable HTTP — add via Tools & Integrations.
Set up →Dust
Enterprise agent builder. Add once in Spaces → Tools; available to all agents.
Set up →Goose
Open-source local AI agent by Block. Add as a remote Streamable HTTP extension.
Set up →Mattermost
Self-hosted team chat. Admin adds MCP server in System Console → Plugins → Agents.
Set up →Gumloop
AI workflow platform. Add as an MCP Server credential in Settings → Credentials.
Set up →Developer Frameworks
LangChain
Python LLM framework. Use langchain-mcp-adapters with transport: http in three lines.
Set up →Langflow
Visual AI app builder. Add via MCP Tools component with HTTP/SSE mode.
Set up →Firebase Studio
Google's Gemini workspace. Configure in .idx/mcp.json; SSE and Streamable HTTP supported.
Set up →