Botstar
BotStar is a comprehensive chatbot platform that enables businesses to design, develop, and train chatbots visually for Messenger and websites.
Verdict
Common use cases
- Script widget open/close behavior for A/B tests
- Log chat events to analytics during support sessions
- Update user profiles mid-conversation from CRM data
- Retrieve webview parameters for debugging modal flows
- Reinitialize widget config after page state changes
Integration
- Vendor
- Botstar
- Category
- communication
- Auth
- API_KEY
- Tools
- 9
- Composio slug
botstar
Tools
- BotStar LiveChat onClose Callback
Tool to register a callback when the chat window is closed. use after the widget is initialized. example prompt: "register an onclose handler that logs 'goodbye!' to the console."
- Close BotStar Livechat Widget
Tool to hide the live chat window. use when the chat widget is configured in livechat or popup mode.
- Get BotStar Application IDs
Tool to retrieve the botstar application id (`appid`). use when initializing or reinitializing the live chat widget.
- Get BotStar Webview Parameter
Tool to retrieve a parameter value passed from the botstar chatbot to the webview. use inside onchatbotready after your page loads in modal mode with bs:input meta tags.
- Livechat boot
Tool to reinitialize the live chat widget with provided data. use after initial load to reset or update widget configuration.
- Livechat on open
Tool to register a callback when the chat window is opened. use after widget initialization.
- Livechat open
Tool to show the live chat window. use after the widget has been bootstrapped with botstarapi('boot') to programmatically open the chat window (mode must be 'livechat' or 'popup').
- Livechat update
Tool to update user details on the current live chat session. use when you need to modify user profile attributes during an active conversation.
- Webview send response
Tool to send data from the webview back to the botstar chatbot. use when you need to transmit responses or custom outputs from an open webview.
Setup
Setup guide
- 11. In Switchy, open your workspace settings and navigate to the MCP integrations page. 2. Click 'Add Integration' and select Botstar from the communication category. 3. You'll be prompted to paste your Botstar API key — find this in your Botstar dashboard under Settings > API Access. 4. Grant the integration access to your workspace; Switchy will confirm the connection succeeded. 5. Open any Space and type '@Botstar get application IDs' to verify the MCP responds with your app ID. 6. To invoke other tools, @mention Botstar followed by a natural-language instruction (e.g. '@Botstar open the live chat widget' or '@Botstar register an onClose callback that logs the session ID'). 7. If a tool requires the widget to be initialized first, Botstar will prompt you to run the boot command before proceeding.
What teammates see: by default, memories from Botstar are scoped to the Space (PROJECT visibility) - you can mark any memory PRIVATE or share it ORG-wide.
Works well with
Top models
Compatibility data appears once enough Spaces have used this MCP together with a given model.
How Switchy teams use it
Starter prompts
Open Chat on Page Load
@Botstar open the live chat widget as soon as the page finishes loading. Make sure the widget is in livechat or popup mode first.Open in a Space →
Log Close Events
@Botstar register an onClose callback that logs 'Chat closed at [timestamp]' to the browser console. Include the session ID if available.Open in a Space →
Update User Details
@Botstar update the current live chat session with the user's email and name. Use email 'visitor@example.com' and name 'Alex Chen' as test values.Open in a Space →
Retrieve App ID
@Botstar get the Botstar application ID for this workspace. Return it as a plain string I can copy into my config file.Open in a Space →
Reinitialize Widget
@Botstar reinitialize the live chat widget with updated user data. Set the language to 'es' and the initial message to 'Hola, ¿cómo puedo ayudarte?'Open in a Space →
Example outputs
Illustrative - representative of the model's voice and quality, not literal recordings.
@Botstar show me the current application ID and open the live chat widget for our support team
Retrieved your Botstar application ID: `app_5x9k2m7n`. The live chat widget is now visible on the page. This example assumes the widget was previously initialized in 'livechat' or 'popup' mode. If the widget hasn't been bootstrapped yet, you'll need to call the boot tool first with your configuration settings.
This demonstrates read + action: fetching the app ID (useful for debugging or re-initialization) and programmatically opening the chat window. Requires the widget to already be configured in livechat or popup mode—won't work if your deployment uses modal-only mode.
@Botstar update the current chat session with user email support@acme.com and name 'Jordan Lee', then register a callback to log when they close the window
Updated the active live chat session with email `support@acme.com` and display name `Jordan Lee`. Registered an onClose callback that will log 'Chat closed by user' to the browser console when the widget is dismissed. These changes apply immediately to the current conversation and persist until the session ends or the page reloads.
Showcases write operations: updating user profile attributes mid-conversation and attaching event handlers. Useful for progressive profiling or analytics tracking. The callback fires client-side only—server-side logging requires additional integration with your backend.
@Botstar retrieve any webview parameters passed from the chatbot, then summarize what data the bot sent to this page
Extracted webview parameters: `user_intent=pricing_inquiry`, `plan_tier=enterprise`, `session_id=sess_8k3m`. Summary: The chatbot routed this user to the pricing page because they expressed interest in enterprise plans during the conversation. The session ID links back to the bot transcript for context. This data was passed via `bs:input` meta tags when the page loaded in modal mode.
Highlights the webview parameter retrieval tool, which bridges bot conversations and embedded web content. Only works when your page is loaded inside a Botstar modal webview—returns empty if accessed directly. Requires the onChatbotReady event to fire before parameters are available.
Use-case deep-dives
When Botstar fits a product-led growth support stack
A 6-person SaaS startup runs onboarding through a chatbot that escalates to live agents when users hit billing or integration questions. Botstar's livechat tools let you programmatically open the widget when a user clicks "Talk to support" in your app, update their profile mid-conversation (passing plan tier or account ID), and register callbacks to log handoff events back to your analytics. This works if your support volume is under 200 conversations a week and you already use Botstar's chatbot platform. If you're stitching together Intercom or Zendesk with a separate bot vendor, the integration overhead outweighs the benefit. The API key auth and 9-tool scope make this a clean fit for teams already committed to Botstar's ecosystem.
Botstar for high-intent visitor engagement at scale
A 12-person B2B marketing team runs paid campaigns to product landing pages and wants to capture leads before they bounce. Botstar's widget can be opened programmatically when a visitor scrolls past the fold or idles for 15 seconds, then update user details as they answer qualification questions in the chat. The onClose callback logs abandonment events to your CRM. This setup works if you're driving 500-plus visitors a day and need chat to feel native to your site, not bolted on. If your traffic is lower or you're already using HubSpot's chat, the integration lift isn't worth it. The real win is when you need the chatbot logic and the live handoff in one vendor, not two.
When Botstar misses the mark for high-touch CS workflows
A 20-person customer success team manages 80 enterprise accounts and wants to automate quarterly check-in outreach via chat. Botstar's livechat tools can open a widget and pass account metadata, but the 9-tool scope doesn't include proactive messaging triggers or CRM sync beyond manual updates. If your CS motion depends on Salesforce data or scheduled outbound nudges, you'll spend more time building workarounds than running check-ins. This MCP is a better fit for reactive support scenarios where the customer initiates the conversation. For proactive enterprise CS, look at integrations that hook directly into your CRM and support workflow automation, not just widget lifecycle events.
Frequently asked
What does the Botstar MCP do in Switchy?
It lets your AI agents control Botstar's live chat widget on your website. Agents can open or close the chat window, register callbacks when users interact with it, retrieve application IDs, and update user profile data during active conversations. Useful if you want AI to orchestrate chat flows based on customer behavior or backend events.
Do I need admin access to connect Botstar?
You need an API key from your Botstar account, which typically requires admin or developer permissions. The MCP doesn't use OAuth, so whoever connects it must have access to generate and copy the key from Botstar's dashboard. Once connected, any Switchy team member with access to the workspace can invoke the tools.
Can the MCP send messages on behalf of the bot?
No. The tools focus on widget lifecycle—opening, closing, registering event handlers, updating user metadata. To send actual bot messages or trigger flows, you'd use Botstar's separate messaging API or configure triggers inside Botstar itself. This MCP is for controlling the chat interface, not authoring conversation content.
Why use this instead of Botstar's JavaScript SDK directly?
If you already have custom frontend code, the SDK is simpler. The MCP shines when you want AI agents to react to backend events—like opening the chat widget when a payment fails or updating user attributes after a CRM sync—without writing glue code. It centralizes control in Switchy's workspace instead of scattering logic across repos.
Who on the team should connect this integration?
Whoever owns your Botstar account and has API key access. Usually a developer or marketing ops person. After connection, your AI agents can invoke the tools autonomously, so make sure the team understands which workflows will trigger widget actions—you don't want the chat popping open unexpectedly during demos.