Chatbotkit

@chatbotkit list all active bots and show me which ones have the most conversations this month

Found 4 active bots in your account. 'Support Bot' leads with 847 conversations, followed by 'Sales Assistant' (312), 'Onboarding Guide' (89), and 'FAQ Helper' (34). Support Bot was created 6 months ago and uses the 'customer-service' skillset. Sales Assistant integrates with Slack and was last updated 3 days ago. This data reflects conversations initiated since December 1st based on the List Bots and List Conversations endpoints.

This example demonstrates read-heavy operations combining bot metadata with conversation counts. Requires API key with read access to both bots and conversations. The MCP paginates through results automatically, but large accounts may hit rate limits when fetching conversation counts for many bots simultaneously.

@chatbotkit create a new skillset called 'product-docs-q1' and add it to our main support bot

Created skillset 'product-docs-q1' with ID sk_abc123. The skillset is now available for assignment but contains no skills yet—you'll need to add abilities or datasets separately. I've updated 'Support Bot' (bot_xyz789) to include this skillset alongside its existing 'customer-service' skillset. The bot will now have access to both skillsets when responding to conversations. Changes take effect immediately for new conversations.

This example shows a write operation that modifies your ChatBotKit configuration. Creating skillsets and updating bot assignments are permanent actions—the MCP cannot undo them. You'll need API key permissions for both skillset creation and bot updates. Consider testing new skillsets on a staging bot before adding to production.

@chatbotkit pull the last 20 messages from conversation conv_456 and draft a summary email for our product team about the customer's feature requests

Retrieved 20 messages from conversation conv_456 spanning a 15-minute support session. The customer (user_789) requested three features: bulk export to CSV (mentioned twice, high priority), dark mode for the dashboard, and Zapier integration for automated workflows. They expressed frustration with manual data entry and praised the current API documentation. Draft email: 'Customer conv_456 strongly advocates for CSV bulk export (urgent pain point), plus dark mode and Zapier support. Positive feedback on API docs suggests they're technical users who would benefit from automation features.'

This example pairs ChatBotKit's conversation retrieval with AI synthesis to extract actionable insights. The MCP fetches raw message data; the AI analyzes tone and priority. Conversation access requires appropriate API permissions, and message history depends on your account's retention settings. Pagination handles conversations longer than 20 messages.

When ChatbotKit wins for small support teams routing tickets

A 5-person SaaS support team fields 200+ inbound questions daily across Slack, Discord, and email. ChatbotKit's integration tooling lets you wire up bots to each channel without custom code—create the integration, point it at a bot, and the MCP handles message routing. The conversation history tools (List Conversations, List conversation messages) mean your team can audit what the bot said before escalating to a human. This works when your knowledge base is stable and under 10k documents; if you're updating docs hourly or need sub-second retrieval across 50k articles, you'll hit pagination lag. If your support load is predictable and you want one AI layer across three channels by Friday, ChatbotKit delivers.

When this MCP helps agencies manage client chatbot fleets

A 3-person agency builds custom bots for 8 clients, each with different tone, skillsets, and data sources. ChatbotKit's List Bots and List Skillsets tools let you programmatically audit which bot uses which skills—critical when a client asks "did we deploy the refund policy update to production?" The Create Skillset tool means you can snapshot a working skill configuration and reuse it across bots without manual copy-paste. This shines when you're managing 5-15 bots and need to diff configurations or roll back a bad deploy. If you're running 50+ bots or need Git-style branching, the pagination and lack of native versioning become friction. For small agencies shipping bots weekly, this MCP turns bot management into a scriptable workflow instead of a dashboard clicking marathon.

When ChatbotKit works for post-launch conversation analysis

A 6-person product team just launched an onboarding bot and wants to see where users drop off or ask questions the bot can't answer. The Complete Conversation and List conversation messages tools let you replay entire threads in Switchy, annotate the failure points, and feed those back into your next skillset iteration. This is faster than exporting CSVs or clicking through a vendor dashboard when you're reviewing 20-30 conversations per sprint. The trade-off: if you need real-time alerting ("notify me when confidence drops below 70%") or want to join conversation data with your analytics warehouse, the MCP doesn't expose webhooks or bulk export. For weekly retro-style conversation reviews with a small team, this tooling gets you from "what happened" to "what to fix" in one Switchy session.