Logo Linear

Linear in Claude and ChatGPT

In short

The Fezola Linear MCP connector relays Linear's official MCP server, so Claude and ChatGPT can read and update issues, projects, cycles, documents and comments. One authorization covers one Linear workspace, and you can connect several.

Connect Linear

Updated on

What your assistant can do

What your assistant can do
ActionRead or writeDetail
IssuesRead and writeSearch, read, create and update issues.
Projects and cyclesReadWhat is in the current cycle, projects due soon.
CommentsRead and writeSummarize a thread; add a comment with your confirmation.
DocumentsReadRead project documents.
Assignees and statesWriteAssign or move an issue, with approval.

The tools are those Linear's server exposes and they can change, so Fezola does not freeze them in a list. Fezola requests read and write permissions, which you see on Linear's screen.

When to choose Fezola

  • Several Linear workspaces: each authorization is one workspace, so several workspaces are several named sources.
  • The same connection in Claude and ChatGPT: one URL serves both. ChatGPT custom connectors are available on Business, Enterprise and Edu plans.
  • Nothing to install: you approve on Linear's own screen and Fezola hosts the connector.
  • One tool, one assistant: Linear's own MCP server can be connected directly in a compatible client. Check its documentation for the current steps.
  • A shared Memory: your preferences follow you across assistants.

Several accounts

Each authorization is one Linear workspace, so several workspaces are several sources, and an answer says which one it comes from. Example: "List blocked issues in both workspaces."

Questions to ask (example prompts)

  • “What is in the current cycle and who owns each issue?”
  • “Create an issue for this bug and assign it to me.”
  • “Summarize the comments on this issue.”
  • “Which issues are blocked?”
  • “List the projects due this month.”

A typical use

A team lead asks: "What is in the current cycle, which issues are blocked, and summarize the comments on the oldest one." The assistant reads the cycle and the thread and reports. It creates or updates nothing unless you ask, and it should confirm first. Example scenario, not a customer story.

Safety

Linear's server asks for read and write permissions, so your assistant can create and update issues: keep its approval prompts on. What Linear returns is treated as data, never as instructions. You can disconnect an account in Fezola at any time and hide Linear from an assistant with the tool ticks.

How to connect it

  1. 1In Sources, add Linear and approve on Linear's screen.
  2. 2Add Fezola to Claude or ChatGPT at the connector URL.
  3. 3Ask about your issues.

Step by step: add a custom connector to Claude, add an MCP server to ChatGPT, FAQ.

Price

Free within your 2 free sources. Pro is unlimited, $9 a month or $90 a year, with a 7-day free trial.

FAQ

Is it Linear's official server?

Fezola relays it.

Can it create and update issues?

Yes: Fezola requests read and write.

Several workspaces?

Yes: one authorization covers one Linear workspace, so several workspaces are several sources in Fezola.

Does it need an API key?

No: you approve on Linear's screen.

Does it work in ChatGPT?

Yes, as a custom connector on Business, Enterprise and Edu plans.

Keep reading

Verified in the code on .