Skip to main content

Linear MCP

The Linear Model Context Protocol (MCP) server gives AI models and agents a standardized way to access your Linear data. Use it from Cequence AI Gateway to let AI assistants find, create, and update issues, projects, comments, and related objects in Linear—with more functionality added over time.

1. Overview​

Linear MCP is a remote, vendor-hosted MCP server provided by Linear. You connect to it from Cequence AI Gateway; no local installation is required.

  • Server URL: https://mcp.linear.app/mcp
  • Transport: HTTP (Streamable HTTP recommended); SSE also supported at https://mcp.linear.app/sse
  • Hosted by: Linear

Linear recommends Streamable HTTP for better reliability. The server follows the authenticated remote MCP spec.

2. Supported authentication types​

TypeSupportedNotes
OAuth 2.1YesRecommended for Cequence AI Gateway. Uses Dynamic Client Registration (DCR); you typically do not need to create an OAuth app in Linear.
API key / Bearer tokenYesFor programmatic or integration use. Use a Linear API key or OAuth access token in the Authorization: Bearer <token> header. Supports read-only access via restricted API keys or acting as an app user.

When you add Linear MCP in Cequence AI Gateway, authentication is typically OAuth 2.1 with DCR. For automation or restricted access, you can use an API key or OAuth token in the Bearer header instead of the interactive OAuth flow.

3. What can you do with this MCP server​

With the Linear MCP server, you can:

  • Find issues — Search and retrieve issues by criteria, status, or project.
  • Create and update issues — Create new issues and update existing ones (e.g. status, assignee, description).
  • Manage projects — Work with Linear projects and their configuration.
  • Handle comments — Add and manage comments on issues.
  • Extend workflows — Use AI to triage issues, draft descriptions, update status, or summarize project progress from your IDE or AI client.

Feedback on functionality is welcomed via Linear support.

4. Prerequisites​

Before adding Linear MCP in Cequence AI Gateway, ensure you have:

  • Access to Cequence AI Gateway (e.g. beta.aigateway.cequence.ai)
  • A Linear account with access to the workspace(s) you want AI tools to use
  • For OAuth authentication: an auth app with client credentials in your vendor account, unless the server supports Dynamic Client Registration (DCR). Linear MCP supports DCR, so you typically do not need to create an OAuth app in Linear.
  • For API key / Bearer: a Linear API key (e.g. from Settings → Account → Security & Access) if you use token-based auth.

5. Example workflows​

  • “List open issues assigned to me in the current sprint.”
  • “Create an issue in project X: title ‘Fix login timeout’, description …”
  • “Update issue LIN-123 to Done and add a comment: ‘Deployed in v2.1.’”
  • “Summarize all issues in project Y that are still in Backlog.”

6. Connecting MCP server from Cequence AI Gateway​

  1. Log in to Cequence AI Gateway.
  2. Choose your tenant.
  3. Go to App catalogue.
  4. Filter by Remote MCP server.
  5. Search for Linear MCP and then select it.
  6. Click Create MCP server.
  7. Choose auth method. If OAuth, you need an auth app with client credentials in your vendor account (see Prerequisites). Linear supports DCR, so OAuth setup is usually straightforward.
  8. Complete the setup as prompted, select tools, and deploy.

Use the generated MCP server URL in your client (Claude, Cursor, Windsurf, etc.) as described in the Client Configuration docs. For detailed UI steps and screenshots, see Create a third-party MCP Server.

7. Additional information​

  • Transports: Streamable HTTP at https://mcp.linear.app/mcp (recommended); SSE at https://mcp.linear.app/sse for clients that do not yet support Streamable HTTP.
  • Troubleshooting: If you see an internal server error when connecting, try clearing saved auth: rm -rf ~/.mcp-auth (where applicable) and reconnect. You may need a newer Node.js version if using mcp-remote.
  • WSL on Windows: If you have issues, some clients suggest using SSE with a wsl wrapper; for Cequence AI Gateway, use the generated URL and auth as configured in the gateway.
  • Official documentation: Linear MCP server.