Team Agent
Team Agent lets you build custom AI agents and deploy them as chatbots in the messaging platforms your team already uses — Slack, Microsoft Teams, Discord, HubSpot, and Zendesk. Connect MCP servers and knowledge bases so your agent can automate real tasks, not just answer questions.
How it works
You create an agent, configure its instructions and capabilities, then deploy it as a chatbot in one or more platforms. Team members interact with the agent directly in their messaging tool — mention it in a Slack channel, assign it a Zendesk ticket, or message it in Teams.
Unlike Inbox Agent, which suggests drafts for you to review, Team Agent acts autonomously. It reads messages, executes tools, queries knowledge bases, and posts responses on its own.
Deploying chatbots
Each agent has a Chatbots tab where you connect it to messaging platforms. One agent can be deployed to multiple platforms simultaneously.
| Platform | How it works |
|---|---|
| Slack | Responds to @mentions and DMs in any connected channel |
| MS Teams | Responds in team channels and group chats |
| Discord | Responds to mentions in your Discord server |
| HubSpot | Assists with conversations in your HubSpot inbox |
| Zendesk | Posts notes or comments on support tickets |
Connecting tools for task automation
Team Agent becomes far more useful when it can take action beyond generating text. By connecting MCP servers, you give your agent access to external tools — querying databases, creating tickets, looking up CRM records, or running API calls.
For example:
- Connect Linear and the agent can create issues or look up project status.
- Connect HubSpot and it can pull customer data to answer sales questions.
- Connect Notion and it can search your internal wiki to ground its responses.
Prerequisites
- A Runbear account (sign up at runbear.io)
- Admin access to the messaging platform where you want to deploy