MCP Server

Connect AI assistants like Claude, Codex, and Cursor to your Conferbot workspace using the Model Context Protocol (MCP). Query chatbots, analyze responses, manage webhooks, and generate reports — all through natural language.

💡

What is MCP?

MCP is an open standard by Anthropic that lets AI assistants interact with external tools and data sources. Instead of copy-pasting API responses, your AI assistant can directly query your Conferbot workspace.

Prerequisites

Starter plan or above — MCP uses the same API access as the REST API
API key — Generate one from your workspace settings under Developer → API Keys
MCP-compatible client — Claude Code, Cursor, Windsurf, JetBrains, OpenAI Codex, or any client supporting the MCP standard

Endpoint

The MCP server uses Streamable HTTP transport (stateless mode) over a single endpoint:

MCP Endpoint

POST https://api-v2.conferbot.com/api/v1/mcp

Authenticate with the x-api-key header, same as the REST API. Each MCP request counts as one API call against your monthly quota.

Connect your IDEs to Conferbot MCP

1Copy the configuration

Claude Code
{
  "mcpServers": {
    "conferbot": {
      "type": "url",
      "url": "https://api-v2.conferbot.com/api/v1/mcp",
      "headers": {
        "x-api-key": "ENTER_TOKEN_HERE"
      }
    }
  }
}

2Add to Claude Code

Add to config file

Create this file in your project root and paste the JSON. If it already exists, add the "conferbot" entry inside the existing mcpServers object:

.mcp.json

Replace the placeholder token

Find the highlighted ENTER_TOKEN_HEREin the configuration above and replace it with your API token. Generate a new token in the section above if you don't have one.

Restart Claude Code

The MCP server will be picked up automatically on restart.

Verify connection

Confirm the server is running by executing this command:

claude mcp list

✓

You're all set!

Once connected, you can manage your Conferbot chatbots using natural language right from your IDE. Try something like:

"List all my chatbots and show their response counts"

Testing with MCP Inspector

Use the official MCP Inspector to test your connection:

Terminal

npx @modelcontextprotocol/inspector

Enter the endpoint URL and add the x-api-key header in the inspector UI.

Available Tools

Tools are actions your AI assistant can invoke. Conferbot exposes 45 tools across chatbots, responses, analytics, webhooks, flows, templates, knowledge base, widget configuration, and conversations.

Chatbot Tools

Tool	Description	Parameters
`list_chatbots`	List all chatbots in your workspace with response counts	None
`get_chatbot`	Get detailed information about a specific chatbot	chatbotId (required)
`create_chatbot`	Create a new chatbot in your workspace	name (required, max 100), description
`update_chatbot`	Update a chatbot's name, description, or disabled status	chatbotId (required), name, description, disabled
`delete_chatbot`	Permanently delete a chatbot from the workspace	chatbotId (required)
`duplicate_chatbot`	Create a copy of a chatbot with all its configuration	chatbotId (required)

Response Tools

Tool	Description	Parameters
`list_responses`	List responses with pagination and date filtering	chatbotId (required), page, limit (max 100), startDate, endDate
`get_response`	Get a single chatbot response by ID	chatbotId (required), responseId (required)

Analytics Tools

Tool	Description	Parameters
`get_chatbot_analytics`	Get analytics including response counts and daily trends	chatbotId (required), days (1–365, default: 30)
`get_api_usage`	Get current month API usage statistics	None

Webhook Tools

Tool	Description	Parameters
`list_webhooks`	List all webhook subscriptions	chatbotId (optional filter)
`get_webhook`	Get details of a specific webhook	webhookId (required)
`create_webhook`	Create a new webhook subscription	chatbotId, url, events[], description
`update_webhook`	Update a webhook (URL, events, status, description)	webhookId (required), url, events, status, description
`delete_webhook`	Delete a webhook and its delivery logs	webhookId (required)
`regenerate_webhook_secret`	Regenerate the signing secret (shown once)	webhookId (required)
`get_webhook_delivery_logs`	Get delivery logs for debugging	webhookId (required), page, limit (max 50)

Flow Tools

Tool	Description	Parameters
`get_chatbot_flow`	Get the conversation flow structure of a chatbot	chatbotId (required)
`update_chatbot_flow`	Replace the entire conversation flow and save a version	chatbotId (required), flow (required), versionName (max 32 chars)
`add_flow_node`	Add a node to the chatbot's conversation flow	chatbotId (required), node (required: id, type, data, position)
`delete_flow_node`	Remove a node and its connected edges from the flow	chatbotId (required), nodeId (required)
`connect_flow_nodes`	Connect two nodes with an edge in the flow	chatbotId (required), edge (required: source, target, sourceHandle, targetHandle, label)
`list_flow_versions`	List saved flow versions for a chatbot	chatbotId (required)
`publish_flow_version`	Publish a specific flow version (or draft) as the live flow	chatbotId (required), versionId (optional, omit to publish draft)

Template Tools

Tool	Description	Parameters
`list_templates`	List available chatbot templates with pagination	category, page, limit (max 50)
`get_template_categories`	Get all available template categories	None
`create_from_template`	Create a new chatbot from a template	templateId (required), name (max 100)

Knowledge Base & AI Tools

Tool	Description	Parameters
`list_knowledge_base`	List knowledge base articles with pagination and filtering	chatbotId (required), categoryId, page, limit (max 50)
`get_knowledge_article`	Get a knowledge base article with full content and metrics	chatbotId (required), articleId (required)
`create_knowledge_article`	Create a new knowledge base article	chatbotId (required), title (required), description (required), content (required), categoryId, seoKeywords[], metaDescription
`update_knowledge_article`	Update a knowledge base article's content or status	chatbotId (required), articleId (required), title, description, content, reviewStatus (draft/in_review/published), seoKeywords[]
`list_knowledge_categories`	List all knowledge base categories for a chatbot	chatbotId (required)
`create_knowledge_category`	Create a new knowledge base category	chatbotId (required), name (required), description, icon (max 10 chars), color (hex)
`get_ai_config`	Get AI agent configuration for a chatbot	chatbotId (required)
`update_ai_config`	Update AI agent settings (mode, greeting, enabled status)	chatbotId (required), enabled, mode (ai_first/human_first/ai_only/human_only), greetingOverride

Widget & Channel Tools

Tool	Description	Parameters
`get_widget_config`	Get the widget appearance and behavior settings	chatbotId (required)
`update_widget_config`	Update widget colors, position, CTA text, and more	chatbotId (required), config (object with headerBgColor, botMsgColor, widgetPosition, chatIconCtaText, enableKnowledgeBase, botAutoOpen, etc.)
`list_channels`	List connected messaging channels for a chatbot	chatbotId (required)
`get_channel_config`	Get configuration for a specific channel connection	channelId (required)
`export_chatbot`	Export a chatbot's full configuration as JSON	chatbotId (required)

Conversation Tools

Tool	Description	Parameters
`list_conversations`	List conversations with filtering by status, channel, and date	chatbotId (required), page, limit (max 50), status, channel, startDate, endDate
`get_conversation`	Get a conversation with all messages and metadata	chatbotId (required), conversationId (required)
`get_conversation_analytics`	Get analytics for a specific conversation (sentiment, duration, etc.)	chatbotId (required), conversationId (required)
`list_tickets`	List support tickets with status filtering	chatbotId (required), status (open/closed/in-progress/pending), page, limit (max 50)
`get_conversation_summary`	Get conversation statistics and trends over a period	chatbotId (required), days (1–365, default: 30)

Resources

Resources are read-only data your AI assistant can access directly. They provide a quick way to fetch workspace data without invoking tools.

Conferbot exposes 6 resources, including static workspace resources and templated chatbot-specific resources.

URI	Description
`conferbot://chatbots`	All chatbots in your workspace with response counts
`conferbot://chatbot/{chatbotId}`	Detailed information about a specific chatbot
`conferbot://chatbot/{chatbotId}/analytics`	30-day analytics data for a specific chatbot
`conferbot://chatbot/{chatbotId}/widget`	Widget appearance configuration for a specific chatbot
`conferbot://chatbot/{chatbotId}/conversations`	Recent conversation summary for a specific chatbot
`conferbot://usage`	Current month API usage statistics

Prompt Templates

Prompts are pre-built workflows that guide your AI assistant through multi-step analyses. Invoke them by name in your AI client.

Prompt	Arguments	What It Does
`analyze-chatbot-performance`	chatbotId (required)	Fetches chatbot details, analytics, and recent responses, then synthesizes a performance analysis with trends and recommendations.
`debug-webhook-deliveries`	webhookId (required)	Fetches webhook config and delivery logs, then analyzes failure patterns and suggests fixes.
`generate-analytics-report`	chatbotId (required), days (optional)	Generates a formatted analytics report with key metrics, daily trends, and actionable insights.
`build-chatbot-flow`	chatbotId (required), description (required)	Guides building a chatbot conversation flow from a natural language description using nodes and edges.
`optimize-chatbot`	chatbotId (required)	Analyzes chatbot performance, flow structure, widget config, and AI settings, then provides actionable improvement suggestions.
`setup-knowledge-base`	chatbotId (required), topic (required)	Guides setting up a knowledge base with categories, articles, and AI configuration for a specific topic.

Example Conversations

Once configured, you can ask your AI assistant questions in natural language. The assistant will automatically invoke the right tools.

List Your Chatbots

You:"Show me all my chatbots and their response counts."

Assistant calls list_chatbots and presents a formatted table.

Analyze Performance

You:"How is my Support Bot performing this month? Any trends I should know about?"

Assistant uses the analyze-chatbot-performance prompt to fetch chatbot details, analytics, and recent responses, then provides a comprehensive analysis.

Debug Webhooks

You:"My webhook seems to be failing. Can you check what's going on?"

Assistant calls list_webhooks to find the failing webhook, then uses get_webhook_delivery_logs to analyze failure patterns and HTTP status codes.

Create a Webhook

You:"Set up a webhook for my Support Bot that sends response.created events to https://hooks.example.com/conferbot."

Assistant calls create_webhook with the specified parameters and returns the signing secret.

Build a Conversation Flow

You:"Build a lead capture flow for my Sales Bot that collects name, email, and company size."

Assistant uses the build-chatbot-flow prompt to design and create flow nodes with add_flow_node and connect them with connect_flow_nodes.

Set Up a Knowledge Base

You:"Set up a knowledge base for my product support chatbot with categories for billing, integrations, and troubleshooting."

Assistant uses the setup-knowledge-base prompt to create categories, articles, and enable AI-powered responses.

Customize Widget Appearance

You:"Change my chatbot's widget header color to dark blue and update the CTA text to 'Need help? Chat with us!'"

Assistant calls update_widget_config to apply the color and CTA text changes.

Review Conversations

You:"Show me a summary of conversations from the past week. Are there any escalated tickets I should review?"

Assistant calls get_conversation_summary and list_tickets to provide a complete overview.

Rate Limits & Usage

MCP requests share the same limits as the REST API:

Rate limit: 60 requests per minute per API key
Monthly quota: Each MCP request counts as one API call

Plan	Monthly API Calls
Starter	10,000
Pro	50,000
Business	200,000
Enterprise	Unlimited

💡

Tip

Use the get_api_usage tool to check your current month usage at any time.

Troubleshooting

401 — Invalid API Key

Your API key is missing, invalid, or has been revoked. Generate a new one from your workspace Developer settings.

403 — Plan Not Supported

Your current plan does not include API access. Upgrade to Starter or above.

429 — Rate Limit Exceeded

You've exceeded 60 requests/minute or your monthly API call quota. Wait a moment and retry, or upgrade your plan for a higher quota.

Connection Timeout

Ensure your firewall or proxy allows outbound HTTPS connections to the API server. The MCP endpoint uses standard HTTPS (port 443).

Tools Not Appearing in Client

Verify the endpoint URL and headers in your client configuration. Test with the MCP Inspector first:

npx @modelcontextprotocol/inspector

Security

Workspace isolation: Your API key only accesses your own workspace data. Cross-workspace access is not possible.
Input validation: All tool inputs are validated with strict schemas before reaching your data.
No secret exposure: Webhook signing secrets are never returned in list or get operations.
Stateless transport: Each request is independent — no session state is stored on the server.
Rate limiting: Per-key rate limiting prevents abuse.

⚠️

Keep Your API Key Secure

Treat your API key like a password. Never commit it to version control or share it publicly. If compromised, revoke it immediately from your workspace Developer settings and generate a new one.