# Dyte MCP for AI Agents AI Agent Connect > Dyte MCP lets you manage live video and audio experiences from any AI agent. It handles meeting creation, participant management, and recording controls for your video infrastructure, allowing you to orchestrate real-time communication workflows through natural language. ## Overview - **Category:** developer-tools - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_1FuAxqsgOTP3wlR6HEGsoXdZYF2SJrn4e2OmQkoI/ai-agent-connect - **Tags:** dyte, video-api, webrtc, live-streaming, meeting-automation ## Description This Connector connects your Dyte organization to an AI agent so you can manage real-time communication workflows without jumping between tabs. Instead of manually clicking through a dashboard to set up a room, you can just tell your agent to spin up a meeting or add a guest. It handles the heavy lifting of generating auth tokens and configuring recording storage, like AWS S3, while you focus on the actual conversation. Because Vinkius hosts this Connector, you get a reliable way to turn your AI into a backend controller for your entire video setup. Imagine you're running a series of technical interviews. Instead of manually creating a new link for every candidate and then checking if the recording started correctly, you can let your agent handle the logistics. You can ask it to see your active sessions to find a specific one or grab a download link for a finished recording without opening a single browser window. It takes the friction out of video management by making your AI the primary point of contact for your infrastructure. Whether you're starting a livestream to a large audience or setting up webhooks to catch meeting events for your internal logs, your agent handles the technical execution. It moves the work from your mouse to your voice, letting you stay in your flow while the agent manages the backend. ## Tools ### add_participant Add a user to a meeting and generate their frontend auth token. This makes it easy to manage guest access without manual intervention. ### create_meeting Create a new Dyte meeting instance instantly. Use this to spin up rooms for interviews or syncs on demand. ### create_webhook Set up real-time notifications for meeting events like starts or ends. This helps sync your internal systems with live data. ### get_recording Fetch the download URL and details for a completed recording. This helps you locate assets quickly after a call. ### list_meetings Fetch a list of all existing meetings in your organization. This is useful for auditing or finding specific sessions. ### list_participants See everyone currently in a specific meeting room. Use this to monitor attendance during live events. ### start_livestream Broadcast a meeting to a wider audience as a live stream. This turns a private call into a public broadcast. ### start_recording Begin a recording with custom storage settings like AWS S3. This ensures your recordings go exactly where you want them. ### stop_livestream End the active live broadcast. Use this to shut down your stream once your event concludes. ### stop_recording End an active recording session. This allows you to control exactly when your video assets are finalized. ### get_meeting Retrieve specific details for a single meeting. Use this to check status or configuration for a specific session. ## Prompt Examples **Prompt:** ``` Create a meeting for a 'Design Review' and tell me the ID. ``` **Response:** ``` I've set up the 'Design Review' meeting. The ID is `dyte_9921`. You can now share this with your team. ``` **Prompt:** ``` Add Alice to the meeting and give me the token. ``` **Response:** ``` Alice has been added. Here is the auth token for your frontend: `eyJhbG...`. ``` **Prompt:** ``` Is the recording for the last meeting finished? Give me the link. ``` **Response:** ``` Yes, the recording for 'Product Sync' is done. You can download it here: `https://s3.amazonaws.com/dyte_rec_123`. ``` ## Capabilities ### Create and manage video meetings Spin up new meeting rooms or fetch details for existing ones instantly. ### Add participants with auth tokens Grant users access to meetings and receive the necessary tokens for your frontend. ### Control recording storage and status Start recordings with specific storage configs and fetch download URLs when finished. ### Start and stop live broadcasts Toggle livestreams to broadcast your meeting content to a wider audience. ### Set up real-time event webhooks Configure notifications to stay updated on meeting starts, ends, and other events. ### See everyone in a session List all current participants in a meeting room to monitor attendance. ## Use Cases ### Webinar automation An ops lead asks the agent to start a livestream and add 50 participants at once to handle a high-traffic event. ### Interview workflows A recruiter has the agent start a recording and fetch the download URL when the call ends to keep the process moving. ### Support calls A dev uses the agent to list meetings and find the right room for a customer issue without navigating a portal. ### Event monitoring A manager uses webhooks to trigger a Slack alert when a high-priority meeting starts to ensure team visibility. ## Benefits - Skip the dashboard by using `create_meeting` to spin up rooms instantly. - Automate guest access with `add_participant` to get auth tokens for your frontend. - Save storage costs by configuring S3 buckets directly through `start_recording`. - Monitor your audience in real-time using `list_participants` to see who's joined. - Get instant updates on meeting status by setting up `create_webhook` for your backend. - Launch broadcasts easily with `start_livestream` and `stop_livestream` commands. ## How It Works The bottom line is you get a hands-off way to control video infrastructure through natural language. 1. Subscribe to the Dyte MCP in the Vinkius catalog. 2. Enter your Dyte Organization ID and API Key in the configuration. 3. Give your AI client commands to manage meetings, recordings, or livestreams. ## Frequently Asked Questions **Can I use Dyte MCP to manage my live streams?** Yes, you can start and stop broadcasts directly through your AI agent. This lets you manage your livestream audience without leaving your current workspace. **How do I add people to a meeting using the Dyte MCP?** The agent handles adding users and gives you the auth token automatically. This simplifies guest access for webinars or private calls. **Does Dyte MCP support saving recordings to my own storage?** Yes, it lets you configure your own storage, like AWS S3, when you start a recording. This ensures your video assets are stored where you need them. **Can I get notified when a meeting starts with Dyte MCP?** You can set up webhooks to send real-time alerts to your own server. This is great for triggering automated workflows or team notifications. **Can the Dyte MCP handle multiple meetings at once?** Yes, it can list, create, and manage many meetings across your organization simultaneously. It gives you a bird's eye view of your entire video infrastructure. **Is Dyte MCP good for automated webinars?** It's perfect for this because it handles participant access and livestreaming via simple commands. You can automate the logistics and focus on the content. **How do I get the authentication token for a new participant to join a meeting?** Use the `add_participant` tool. You need to provide the `meeting_id`, `name`, and a `preset_name`. The tool will return the `authToken` required by Dyte's frontend SDKs. **Can I configure where my meeting recordings are stored?** Yes. When using the `start_recording` tool, you can optionally provide a `storage_config` JSON object (e.g., for AWS S3) to define your custom storage destination. **How can I monitor when a meeting starts or ends in real-time?** You can use the `create_webhook` tool to register a URL that Dyte will notify when specific events occur, such as `meeting.started` or `meeting.ended`.