chatbot6 MCP server
Manage all your chatbots from ChatGPT, Claude, Cursor, VS Code or any MCP client. The AI can list, create and edit your chatbots, update their knowledge, publish them, write blog posts and read your numbers.
Your key
Sign in to see your key. A key is created automatically for every account.
Sign inServer address
Clients that can send headers use this address with the header Authorization: Bearer <key>.
Address with key
For apps that cannot send headers (ChatGPT, Claude.ai). The key is inside the address: keep it secret like a password.
Set up step by step
Pick your app and follow the steps. It takes about a minute.
Claude (claude.ai and Claude Desktop)
- Open Claude and go to Settings, then Connectors.
- Click Add custom connector.
- Enter chatbot6 as the name and paste the address with key below as the server URL. Leave the OAuth fields empty.
- Click Add, then Connect if Claude asks.
- In a new chat, open the tools menu (the sliders icon) and turn chatbot6 on.
- Ask something, for example: show my chatbots and their status.
Claude Code
- Run this command in your terminal:
claude mcp add --transport http chatbot6 https://chatbot6.com/api/v1/mcp \ --header "Authorization: Bearer your_key"
- Start Claude Code and type /mcp to check that chatbot6 is connected.
- Ask in plain words; Claude Code calls the chatbot6 tools when needed.
ChatGPT
- Open ChatGPT on the web and go to Settings, then Apps and Connectors, then Advanced settings, and turn on Developer mode.
- Go back to Apps and Connectors and click Create.
- Enter chatbot6 as the name and paste the address with key below as the MCP server URL.
- Choose No authentication, confirm that you trust this app and click Create.
- In a new chat, click the plus button, choose Developer mode and select chatbot6.
- Ask something; ChatGPT asks for your approval before actions that change data.
Cursor
- Open Cursor Settings, then MCP, then Add new MCP server. You can also edit ~/.cursor/mcp.json directly.
- Paste this configuration and save:
{ "mcpServers": { "chatbot6": { "url": "https://chatbot6.com/api/v1/mcp", "headers": { "Authorization": "Bearer your_key" } } } } - Check that chatbot6 shows a green dot, then use it in Agent mode.
VS Code (GitHub Copilot)
- Create the file .vscode/mcp.json in your project, or run the command MCP: Add Server.
- Paste this configuration and save:
{ "servers": { "chatbot6": { "type": "http", "url": "https://chatbot6.com/api/v1/mcp", "headers": { "Authorization": "Bearer your_key" } } } } - Click Start above the server, open Copilot Chat in Agent mode and enable the chatbot6 tools.
Windsurf, Gemini CLI and other clients
- Any client that supports remote MCP over Streamable HTTP works: use the server address and the header Authorization: Bearer <key>.
- If your client only runs local (stdio) servers, such as older Claude Desktop versions, add this configuration. It uses mcp-remote and needs Node.js:
{ "mcpServers": { "chatbot6": { "command": "npx", "args": [ "-y", "mcp-remote", "https://chatbot6.com/api/v1/mcp?key=your_key" ] } } } - Restart the client and look for the chatbot6 tools.
Tools
What the AI can do through this server. It asks before publishing; chatbots can only be deleted in the studio.
list_chatbotsList the user's chatbots: id, name, status, subscribers, links to the live site and its admin panel.
get_chatbotFull details of one chatbot: texts, instructions, model, plans, numbers of the last 30 days and links.
update_chatbotChange texts and settings of a chatbot. Send only the fields to change; the rest stays. starters and features: one item per line.
set_knowledgeReplace or extend the knowledge text the chatbot answers from (FAQ, product info, prices, policies).
publish_chatbotPublish (live=true) or take back to draft (live=false). Publishing runs an automatic review; the result may be "review".
create_chatbotCreate a new draft chatbot with its own site. Returns its id and links.
list_postsBlog posts of a chatbot site, newest first, with their status (draft, review, queued, writing, ready, failed). Bodies are omitted.
write_postPublish a blog post you wrote on the chatbot site, or replace one (postId). It is checked automatically before going live; status "review" means it waits for a decision.
generate_postLet the AI write a blog post (text and image) about a topic on the chatbot site. It is queued and written in the background with your AI balance; check list_posts for its status.
delete_postDelete a blog post permanently.
list_foldersThe user's studio folders (private grouping of chatbots): id and name. list_chatbots gives each chatbot's folderId.
create_folderCreate a studio folder. Move chatbots into it with update_chatbot folderId.
creator_statsThe user's creator numbers for the last 30 days across all chatbots: active subscribers, new, cancelled, retention, messages, sales.
Writing your own code? Use the REST API. Each chatbot also has its own API and MCP server for its users (chat and blog). You find them on the developers page of the chatbot site.