Think inside your AI world.
Linear MCP: what it is and what it reads
The spec behind one of the most-searched MCP phrases: official, hosted, OAuth 2.1, and a read surface that covers the whole delivery picture.
Linear MCP is Linear's official hosted Model Context Protocol server. Auth: OAuth 2.1. Hosting: by Linear. Surface: issues with priority, labels, assignees and linked PRs; projects; cycles; milestones; initiatives; project updates and health; documents; roadmaps; releases. Clients: anything that speaks MCP — the 'linear mcp claude' search resolves to adding this server in Claude's connectors, and the same server serves ChatGPT, Cursor and the rest.
Why this server reads so well
Linear's data model is unusually disciplined, and the server inherits it: issues arrive with their PRs attached, projects with their health, milestones with their state. Delivery questions that span artefacts — is the launch on track, what does the cycle actually contain — resolve without joins you have to perform in your head.
The tracker tracks; it does not promise
What no board carries: your team's promises about the board. The scope ceiling that makes a cycle honest, the triage clock on a P1, the definition of launch-ready — each is a ratified agreement with an incident behind it, and each lives with the team, not the tool.
Through Unl those agreements read along: every board read arrives with the relevant promise applied, its reasoning attached. The server supplies the state; your standing supplies the verdict.
Linear MCP serves the workspace state, cleanly and officially; the promises you made about that state live with you — and through Unl they are present at every read.
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
Is Linear's MCP server official and hosted?
Yes — Linear builds and hosts it, with OAuth 2.1 authentication. No self-hosting or credential handling on your side.
What does 'linear mcp claude' mean in practice?
Adding this server as a connector in Claude — a one-time OAuth flow. The same server connects to any MCP client; the search phrase just names the commonest pairing.
Where are the setup steps?
On the connect pages linked below — Claude and ChatGPT each get the step-by-step; this page is the spec they share.
What if the workspace is restructured under your promises?
Then the promises survive the restructure, because they were never stored inside it. Teams get renamed and boards get merged; a commitment about how quickly something is answered does not depend on which board it sits on. That independence is the reason to hold it separately.
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.