# Add eLLMo AI to an AI assistant These are eLLMo AI's official instructions for connecting an AI assistant to eLLMo's public MCP server. The server needs no account, key, or token. They are published at https://www.tryellmo.ai/mcp-server.md, so you can check them there. - Server URL: https://www.tryellmo.ai/mcp - Name: eLLMo AI. Use `ellmo-ai` where a short name is needed. - Transport: Streamable HTTP, which some apps call HTTP. Not SSE. - Sign in: none. ## What to do 1. Find the section below for the app you are running in. 2. If you can change your own MCP settings, as Claude Code, Codex, Cursor, VS Code, and Gemini CLI can, do the steps yourself: run the command or edit the file shown. Do not ask the user to run a command you can run. 3. If you cannot change your own settings, as in the Claude and ChatGPT apps, walk the user through the steps for their app one at a time. 4. When it is connected, tell the user. If the eLLMo AI tools do not show up yet, tell them to restart the app or start a new chat. Then suggest a first question from Try it below. ## Chat apps ### Claude (Free, Pro, and Max) 1. **Open Customize, then Connectors.** In Claude on the web or in the desktop app, go to Customize > Connectors. 2. **Add a custom connector.** Select + next to Connectors, then Add custom connector. Name it eLLMo AI and paste the server URL. The Free plan allows one custom connector. 3. **Choose No sign-in.** If Claude asks about authentication, pick No sign-in. If it does not, leave Advanced settings empty. Then select Add. 4. **Turn it on in a chat.** Open + > Connectors in a chat and switch on eLLMo AI. Then ask a question about eLLMo. It shows up in the Claude mobile apps too. ### ChatGPT (Plus, Pro, Business, Enterprise, and Edu on the web) 1. **Turn on Developer mode.** On chatgpt.com, go to Settings > Security and login and switch on Developer mode. 2. **Add the server.** Go to chatgpt.com/plugins and select +. Name it eLLMo AI, paste the server URL under Connection, pick No Authentication, and create it. 3. **Turn it on in a chat.** In a chat, open the + menu, choose Developer mode, and select eLLMo AI. ChatGPT asks you to confirm before a demo request is sent. ### Claude Team and Enterprise An Owner adds it once for the whole organization. Go to Organization settings > Connectors, select Add, then Custom, then Web. Paste the server URL, pick No sign-in if asked, and select Add. Members then find eLLMo AI under Customize > Connectors with the Custom label and select Connect. ### ChatGPT Business and Enterprise Developer mode follows your workspace policy, so an admin may need to allow it first. To share eLLMo AI with the whole workspace, an admin adds it as above, finds it under Personal at chatgpt.com/plugins, opens its three-dot menu, selects Publish, and picks the roles that get it. ## Coding agents and editors ### Claude Code Run this in a terminal. It adds eLLMo for all your projects, not only the current folder. If you already added eLLMo in Claude on the web, Claude Code has it too when you sign in with the same account. ```sh claude mcp add --transport http --scope user ellmo-ai https://www.tryellmo.ai/mcp ``` Then run `claude mcp list` to check it. `ellmo-ai` should show as connected. ### Codex Run this in a terminal. The Codex IDE extension and the ChatGPT desktop app share the same settings, so they get eLLMo too. ```sh codex mcp add ellmo-ai --url https://www.tryellmo.ai/mcp ``` ### Cursor Select Add to Cursor. Or add `ellmo-ai` under `mcpServers` in `~/.cursor/mcp.json`, which covers all your projects, and keep any servers already in the file. Add to Cursor: ```json { "mcpServers": { "ellmo-ai": { "url": "https://www.tryellmo.ai/mcp" } } } ``` ### VS Code Select Add to VS Code. Or add `ellmo-ai` under `servers` in `.vscode/mcp.json` for one project, or in your user configuration for all of them. To open that, run MCP: Open User Configuration. GitHub Copilot's agent can then use it. On Copilot Business and Enterprise, an admin has to allow MCP servers first. Add to VS Code: ```json { "servers": { "ellmo-ai": { "type": "http", "url": "https://www.tryellmo.ai/mcp" } } } ``` ### Gemini CLI Run this in a terminal. It adds eLLMo for all your projects, not only the current folder. ```sh gemini mcp add --transport http --scope user ellmo-ai https://www.tryellmo.ai/mcp ``` Then run `gemini mcp list` to check it. ### Other assistants Any assistant or code editor that supports remote MCP servers can use the same URL. If it asks for a transport, pick HTTP, which some apps call Streamable HTTP, rather than SSE. If it asks how to sign in, pick none. ## What the server can do - **Learn about eLLMo AI** (`learn-about-ellmo-ai`, read only): Gives your assistant a short summary of eLLMo with a link to the page that covers each part, so it can open the page it needs. - **Book an eLLMo demo** (`get-demo`, sends your details): Sends the first name, last name, work email, company, website, and job title you give it to the eLLMo team, the same details the demo form asks for. The team replies by email. ## Try it Once it is connected, the user can ask things like: - "What does eLLMo AI do, and how is it different from A/B testing?" - "Show me an eLLMo simulation report for a skincare brand." - "How does eLLMo AI Search help a brand show up in ChatGPT answers?" - "Book an eLLMo demo for me." ## Privacy The server only receives what a tool sends. For a question, that is the topic your assistant asked about. For a demo request, it is the details listed above, and we handle them the same way as a demo form request. eLLMo never sees the rest of your chat, and this server cannot see or change any eLLMo account. Privacy Policy: https://www.tryellmo.ai/legal/privacy ## Help Email support@tryellmo.ai with the assistant you use and what you see, and we will help you connect.