Six New Integrations Are Live: Shopify, Notion, Linear, Tettra, Figma, and Google Chat

Six new integrations are live in AgentTeams: Shopify, Notion, Linear, Tettra, Figma, and Google Chat. That takes your AI agents from working across a handful of tools to working across the store, the wiki, the issue tracker, the design files, and one more chat surface, all with the same agent you already set up. This is a roundup of what each one does, plus the two things that make these more than connectors.

If you are new here, the model is simple: you hire an AI agent like an employee, connect the tools it should work in, and it handles the work where your team already works. Every integration below ships dedicated tools for the common jobs plus a generic API passthrough, so an agent is never blocked waiting for us to wrap one more action.

What is new, tool by tool

• Shopify. Read and triage orders, look up customers and their history, check and update inventory, and watch for orders that need a human eye (high-value first purchase, odd shipping address, unusual quantity). The e-commerce workflows get their own deep dive in AI agents for e-commerce.
• Notion. Read and write pages, query databases, and pull from your team docs so an agent answers from your actual source of truth instead of guessing.
• Linear. Pick up assigned issues, acknowledge them, read context and comments, and respond when someone asks a direct question, so issues do not sit cold.
• Tettra. Search and read your Tettra knowledge base so support and ops agents answer from approved, written-down knowledge rather than improvising.
• Figma. Read design files, render frames as images so the agent can see what changed, and post comments on behalf of the team when a design needs feedback or a question has a clear answer.
• Google Chat. Read and send messages in spaces and reply where it is tagged, so teams that live in Google Chat get the same responsiveness as teams on Slack or Teams.

What makes these more than connectors

A connector lets an agent call a tool. Two things turn that into an actual coworker, and both apply to every integration above.

• Shared, cross-channel memory. Your agents now remember what they learn and recall it across tools and channels. A customer who explained a problem in a Shopify order note and then followed up in Google Chat is the same person to the agent, not two cold starts. Knowledge an agent picks up doing one task carries into the next, instead of evaporating at the end of a conversation.
• Reactions to events. Agents do not only respond when spoken to. They react when something happens: a new Shopify order, a new Linear issue assigned to them, an @mention in Google Chat, a comment on a Figma file. You decide which events an agent reacts to, and anything customer-facing waits for a human to approve before it sends.

This is also the argument for breadth over depth, which we make in full in horizontal vs vertical AI agents: an agent that spans your tools and remembers across them beats six separate bots that each know one app and forget everything else.

What is coming next

A few more are in the queue and marked coming soon in the catalog while we finish them:

• Zendesk. Read, reply to, and triage tickets, with drafts held for human approval. The connector works; we are finishing the global sign-in path so any Zendesk account can connect cleanly.
• Pipedrive. Read and update deals, contacts, and activities, and move deals through the pipeline.
• vTiger. CRM records, contacts, and pipeline for teams on vTiger.

Where this fits

These six bring the AgentTeams catalog past two dozen live tools, alongside Slack, Google Workspace, Microsoft 365, GitHub, and Help Scout. The point was never one more logo. It was that the agent you hired can now do more of an actual job end to end, because the job rarely lives in a single app. You can browse the full list on the integrations page, and if you are weighing your options, see how the platforms compare.

Ready to connect one? Open your agent's Tools tab and pick the integration you want, or, if you are just starting out, onboard your first agent in under an hour.