Think inside your AI world.
How do I connect Google Sheets to Claude?
Sheets over MCP reads the actual cells — values by A1 range, formulas, tabs, metadata — which matters because most small companies' real model lives in a spreadsheet.
Direct answer: connect Google Sheets over MCP in Claude — Settings → Connectors, authorise your Google account ('claude mcp add' in Claude Code). Claude reads cell values by A1 range, formulas, tab structure and metadata. Given that 23% of startups with a go-to-market motion run their CRM in Sheets, this connector is quietly one of the most consequential on the list.
The spreadsheet, finally conversational
What does the runway tab say now, which rows changed this week, what formula drives the total that looks wrong — read from the live sheet, formulas included. The formula read is underrated: half of spreadsheet debugging is discovering what a cell actually computes.
A model is assumptions wearing numbers
Worked case: your financial model's CAC assumption was ratified — £180 blended, revisited quarterly, because the model's hiring plan collapses if CAC drifts and nobody notices. The sheet computes everything downstream of the assumption flawlessly; that the assumption has an owner, a review date and a fragility story is context about the model no cell contains.
Cells read against their assumptions
Through Unl the model read arrives annotated by your rulings: actuals now imply CAC near £210 — past the assumption you ratified in April, which the hiring plan leans on; the quarterly review is due early. Drift becomes visible at read time, not at the quarter's post-mortem.
Sheets to Claude reads the cells and the formulas; Sheets through Unl reads them against the assumptions you ratified — the model's fragile numbers get watched by the rules that know why they matter.
Reads through Unl arrive with measured context — in the presence of the decisions you’ve already settled. The reach lane is live: one box, paste anything. If it speaks MCP, Unl can reach it. Readings arrive unprompted, the data beside the criterion; Unl is a courier, not a warehouse, and keeps only your keys and the frame.
Read further
Questions people ask
What exactly can Claude read from a connected Sheet?
Cell values by A1 range, formulas, tab structure and metadata — the working substance of the spreadsheet, not an export of it.
Does this connection work for ChatGPT?
Yes — the Sheets MCP surface is standard protocol; other clients connect through their own menus with the same Google authorisation.
Why ratify assumptions instead of commenting them in the sheet?
A cell comment waits to be hovered over. A ratified assumption arrives during the read, carrying its owner, its review date and the reason it is fragile — which is what makes drift a flagged event instead of a discovered one.
What if the assumption only exists inside a formula?
Then nothing can read it, and that is the gap rather than a limitation of the connection. A number hard-coded mid-formula is invisible to everyone including the person who put it there six months ago; naming it outside the cell is what makes it something a rule can check.
What this is
Think inside your AI world — you stay in command
Save the thoughts, decisions and targets worth keeping, each with its reasoning, carried into every AI session the moment they matter. A new unit of exchange between you and your AI: the Settled Why with standing that travels. Unprompted.
Your whole AI world. What you decided at the epicentre. It plugs into Claude, Claude Code, ChatGPT and Cursor as an MCP connector — quick to connect, in a couple of steps.
MCP native·Human settled·Model agnostic·Your data
Measured Context
Connect Unl to bring the right information into the moment.
Your sources, read against the criteria you set.
The full product, open. Free at launch.