Which chats can connect?
Claude today. The connection uses MCP, the Model Context Protocol, an open standard for connecting AI tools, so the door is a protocol rather than a private plug.
Nativeloop connects to Claude, so you can ask about the business, start a routine, and approve the result from the chat you already have open. Nothing lands in your Google files until a person approves the exact change.
Ask, run, and approve from chat. Google-file changes still wait for a person’s approval.
Connecting Claude uses MCP, the Model Context Protocol: the open protocol for connecting AI tools. The connection carries the same permissions as the rest of Nativeloop, and chat becomes one more door into the same instrument: the same record, the same approvals, the same cap on spend. The Security page is specific about what it can reach and how to disconnect it.
The whole team does not need Claude. The app, your sheets, and your inbox carry the same work, and the record reads the same wherever the work was asked for. Connecting chat is one person’s choice, not a migration.
Claude today. The connection uses MCP, the Model Context Protocol, an open standard for connecting AI tools, so the door is a protocol rather than a private plug.
It can propose. A person approves the exact change before it lands, from chat or from the app, and Google-file changes today always wait for that approval. The record keeps what ran, what changed, and who approved it.
Only what you already shared with Nativeloop: the folders, links, and single files you granted. Connecting Claude grants nothing new, and disconnecting it ends new reads and writes from chat.
No. It is included in both plans. Runs started from chat are metered like any other run, in dollars you can read, against the same monthly cap.
We onboard in small batches. A Loop is a routine the company hands off, and we work with each early company to get the first one right. Tell us which routine takes the most time. We will email when there is a fit and a place opens.
Join the waitlist