Who it is for, and what it does well
Claude for teams
Teams that want Anthropic’s assistant with shared projects and admin controls.
- Anthropic’s models and features. The people who build the model build the assistant.
- Nothing to run. Sign up and invite the team.
- Projects and connectors. Shared context and connections to common apps.
Vyre
The agentic command center for AI-native teams: people and agents work in one Space on a server you own.
Industry-neutral. You shape the records, and Flows, Playbooks and a Gate run the work.
Works with Vyre.
How Vyre connects to Claude for teams today.
Today
Anthropic publishes an API, and Vyre works with Claude through your own accounts or keys. A person adds a Connection in the Vault that declares the host, how the key is sent and which operations read, draft, change, send, spend or delete. A Flow then calls the service from a web step, or a custom module wraps it.
Where the vendor publishes an MCP server, a person can add it for agents to use.
Coming in 0.5.2
A packaged Integration, so the Connection and its operations arrive ready-made.
Vyre can be extended by integrating your Zoho, QuickBooks, RingCentral and the rest of your software stack*, Claude for teams included.
*Where the vendor provides an API.
Where Vyre fits.
What Vyre does around it.
Vyre does not replace Claude for teams. These are the parts of the work Vyre does around it.
Chat
Chats where people and agents from any model take part together.
Agents
Agents with a role and a reach, taking tasks.
Records
The records, tasks and files the chat is about, in the same Space.
Vyre runs Claude through your own accounts. Keep the Claude workspace for individual thinking and let Vyre hold the team’s shared work.
Why that helps.
What a team gets from it.
Claude does the thinking, and Vyre gives it the records, the task and the approval.
- One place for people and agentsPeople and agents are members of the same Space. They share the records, tasks, files and chat, so nobody carries data from one tool to the next.
- Agents do the workAn agent takes the task and finishes it, hands it on or marks it stuck with a reason. A person is not moving data between tools by hand.
- Every outward action approvedAn email, a payment, a deletion or a published page waits at the Gate until a person has read the exact words and said yes.
- Data on your own serverRecords, files and the history of every decision stay on a server you own. Sealed fields show an agent a placeholder.
Side by side
| Question | Claude for teams | Vyre |
|---|---|---|
| What it is | A shared Claude workspace for a team. | A shared working layer for people and agents: records, tasks, Flows, documents, sites and a Gate. |
| Where it runs | Hosted by the vendor. | On a server you own. |
| Connecting to the other | Anthropic publishes an API, and Vyre works with Claude through your own accounts or keys. | A Connection in the Vault and a Flow web step or custom module. Coming in 0.5.2: a packaged Integration. |
| Parts of the job in Vyre | Set by the product. | Chat, Agents and Records. |
| Outward actions | Set by how the product is configured. | Held at the Gate until a person approves the exact words. |
The left column comes from the product’s public documentation or describes the kind of product. Where it differs, the vendor’s own documentation is the one to trust.
Questions people ask
Does Vyre replace Claude for teams?
No. Vyre runs Claude through your own accounts. Keep the Claude workspace for individual thinking and let Vyre hold the team’s shared work.
How does Vyre work with Claude for teams?
Anthropic publishes an API, and Vyre works with Claude through your own accounts or keys. In Vyre, a person adds a Connection in the Vault that declares the host, how the key is sent and which operations read, draft, change, send, spend or delete. A Flow then calls the service from a web step, or a custom module wraps it. Coming in 0.5.2: a packaged Integration, so the Connection and its operations arrive ready-made. Integrations work where the vendor provides an API.