Linear MCP Server: Give Your AI Agent Access to Issues, Cycles, and Projects

Why Connect Your AI Agent to Linear

Linear is where engineering teams track work. Issues, cycles, projects, roadmaps, triage queues. An AI agent with Linear access can file bugs from code reviews, update issue status after deploys, pull cycle reports, create issues from Slack conversations, and keep your backlog groomed without manual effort.

Linear has a GraphQL API but no official MCP server. Building a custom integration means learning their schema, managing OAuth, and deploying infrastructure. DataFaucet captures the API from a browsing session and hosts it for you.

Building a Linear MCP Server with DataFaucet

Step 1: Scan Linear

Go to DataFaucet and enter your workspace URL:

https://linear.app/your-team/issues

Step 2: Interact With Key Features

Browse naturally:

• View your issues list (captures issue query)
• Create a new issue (captures mutation with all fields)
• Search for something (captures search endpoint)
• Open a cycle view (captures cycle issues)
• Move an issue between states (captures state update)
• View a project (captures project detail)
• Filter by assignee/label (captures filtered queries)

Step 3: Deploy

DataFaucet generates typed tools:

• search_issues — full-text search across issues
• get_issue — issue details (description, assignee, state, labels, comments)
• create_issue — new issue with title, description, team, assignee, priority, labels
• update_issue — change state, priority, assignee, add labels
• list_cycle_issues — all issues in the current cycle
• get_project — project details and progress
• add_comment — post comments on issues

Connecting to Your Agent

Claude Code:

claude mcp add linear-tools https://datafaucet.dev/api/mcp/your-server-id

Cursor / Claude Desktop:

{
  "mcpServers": {
    "linear": {
      "url": "https://datafaucet.dev/api/mcp/your-server-id"
    }
  }
}

What You Can Build

With Linear MCP tools your agent can:

• Bug filer — create issues directly from error logs or code review feedback
• Cycle reporter — summarize cycle progress, flag at-risk issues
• Deploy linker — mark related issues as done when a PR merges
• Triage assistant — read new issues, suggest priority and assignee based on content
• Standup prep — pull your in-progress issues and blocked items

Linear's API

Linear uses GraphQL at https://api.linear.app/graphql. DataFaucet captures the queries and mutations the web app makes:

| Action | GraphQL Operation |

|--------|-----------------|

| List issues | IssueConnection query with filters |

| Get issue | Issue query by identifier |

| Create issue | issueCreate mutation |

| Update issue | issueUpdate mutation |

| Search | searchIssues query |

| Cycle issues | Cycle.issues connection |

| Add comment | commentCreate mutation |

Auth

Linear web sessions use a cookie-based auth token. DataFaucet captures this during scanning. Sessions last weeks unless you explicitly log out.

For longer-lived access, you can also create a Personal API Key in Linear Settings > API and use it via the cURL input method in DataFaucet.

Linear vs Jira

Both are project management tools, but they target different audiences:

• Linear — fast, keyboard-driven, built for engineering teams
• Jira — enterprise, configurable, built for cross-functional teams

If you use Linear, see this post. If you use Jira, see our Jira MCP server guide.

Tips

• Create an issue during the scan to capture the full field schema (priority, labels, assignee selectors)
• Search in Linear to capture the search query (most useful tool for agents)
• Linear's API is fast (sub-100ms responses) so DataFaucet tools respond quickly
• For team use, scan with a service account or shared workspace admin

Related

• How to create an MCP server — general guide
• Linear use case — quick overview
• Jira MCP server — alternative for Jira teams
• GitHub MCP server — pair with Linear for full dev workflow