Linear MCP Server

A Model Context Protocol (MCP) server implementation that provides access to Linear’s issue tracking system through a standardized interface.

Features

• Create new issues and subissues with label support
• Retrieve the list of linear projects
• Retrieve the project updates
• Create a new project update with health status
• Update existing issues with full field modification
• Delete issue with validation
• Self-assign issues using ‘me’ keyword
• Advanced search with Linear’s powerful filtering capabilities
• Filter issues by cycle (current, next, previous, or specific cycle by UUID or number)
• Add comments to issues with markdown support
• Query Linear issues by ID or key with optional relationships
• Search issues using custom queries with enhanced metadata
• Type-safe operations using Linear’s official SDK
• Comprehensive error handling
• Rate limit handling
• Clean data transformation
• Parent/child relationship tracking with team inheritance
• Label management and synchronization

Prerequisites

• Bun runtime (v1.0.0 or higher)
• Linear account with API access

Environment Variables

LINEAR_API_KEY=your_api_key  # Your Linear API token

Installation & Setup

1. Clone the repository:

git clone [repository-url]
cd linear-mcp

2. Install dependencies and build:

bun install
bun run build

3. Configure the MCP server:

Edit the appropriate configuration file:

macOS:

• Cline: ~/Library/Application Support/Code/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json
• Claude Desktop: ~/Library/Application Support/Claude/claude_desktop_config.json

Windows:

• Cline: %APPDATA%CodeUserglobalStoragesaoudrizwan.claude-devsettingscline_mcp_settings.json
• Claude Desktop: %APPDATA%Claude Desktopclaude_desktop_config.json

Linux:

• Cline: ~/.config/Code/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json
• Claude Desktop: sadly doesn’t exist yet

Add the following configuration under the mcpServers object:

{
  "mcpServers": {
    "linear": {
      "command": "node",
      "args": ["/absolute/path/to/linear-mcp/build/index.js"],
      "env": {
        "LINEAR_API_KEY": "your_api_key"
      }
    }
  }
}

4. Restart the MCP server.

Within Cline’s MCP settings, restart the MCP server. Restart Claude Desktop to load the new MCP server.

Development

Run development server:

bun run dev

Build project:

bun run build

Available MCP Tools

For detailed usage examples of all tools, see USAGE.md.

create_issue

Create a new Linear issue or subissue.

Input Schema:

{
  "teamId": "string",     
  "title": "string",      
  "description": "string",
  "parentId": "string",   
  "status": "string",
  "priority": "number",   
  "assigneeId": "string | 'me'",
  "labelIds": ["string"]  
}

update_issue

Update an existing Linear issue.

Input Schema:

{
  "issueId": "string",    
  "title": "string",
  "description": "string",
  "status": "string",     // Expects status NAME (e.g., "In Progress"). Must be valid for the issue's team.
  "priority": "number",   // Expects 0 (None) to 4 (Low).
  "assigneeId": "string | 'me'",
  "labelIds": ["string"],
  "cycleId": "string"
}

get_issue

Get detailed information about a specific Linear issue with optional relationships.

Input Schema:

{
  "issueId": "string",
  "includeRelationships": "boolean"  
}

search_issues

Search for Linear issues using a query string and advanced filters. Supports Linear’s powerful filtering capabilities.

Input Schema:

{
  "query": "string",
  "includeRelationships": "boolean",
  "filter": {
    "title": { "contains": "string", "eq": "string", ... },
    "description": { "contains": "string", "eq": "string", ... },
    "priority": { "gte": "number", "lt": "number", ... },
    "estimate": { "eq": "number", "in": ["number"], ... },
    "dueDate": { "lt": "string", "gt": "string", ... },
    "createdAt": { "gt": "P2W", "lt": "2024-01-01", ... },
    "updatedAt": { "gt": "P1M", ... },
    "completedAt": { "null": true, ... },
    "assignee": { "id": { "eq": "string" }, "name": { "contains": "string" } },
    "creator": { "id": { "eq": "string" }, "name": { "contains": "string" } },
    "team": { "id": { "eq": "string" }, "key": { "eq": "string" } },
    "state": { "type": { "eq": "started" }, "name": { "eq": "string" } },
    "labels": { "name": { "in": ["string"] }, "every": { "name": { "eq": "string" } } },
    "project": { "id": { "eq": "string" }, "name": { "contains": "string" } },
    "and": [{ /* filters */ }],
    "or": [{ /* filters */ }],
    "assignedTo": "string | 'me'",
    "createdBy": "string | 'me'"
  },
  "projectId": "string",
  "projectName": "string"
}

Supported Comparators:

• String fields: eq, neq, in, nin, contains, startsWith, endsWith (plus case-insensitive variants)
• Number fields: eq, neq, lt, lte, gt, gte, in, nin
• Date fields: eq, neq, lt, lte, gt, gte (supports ISO 8601 durations)

get_teams

Get a list of Linear teams with optional name/key filtering.

Input Schema:

{
  "nameFilter": "string"  
}

delete_issue

Delete an existing Linear issue.

Input Schema:

{
  "issueId": "string"
}

create_comment

Create a new comment on a Linear issue.

Input Schema:

{
  "issueId": "string",
  "body": "string"
}

get_projects

Get a list of Linear projects with optional name filtering and pagination.

Input Schema:

{
  "nameFilter": "string",
  "includeArchived": "boolean",
  "first": "number",
  "after": "string"
}

get_project_updates

Get project updates for a given project ID with optional filtering parameters.

Input Schema:

{
  "projectId": "string",
  "includeArchived": "boolean",
  "first": "number",
  "after": "string",
  "createdAfter": "string",
  "createdBefore": "string",
  "userId": "string | 'me'",
  "health": "string"
}

create_project_update

Create a new update for a Linear project.

Input Schema:

{
  "projectId": "string",
  "body": "string",
  "health": "onTrack | atRisk | offTrack",
  "isDiffHidden": "boolean"
}

Technical Details

• Built with TypeScript in strict mode
• Uses Linear’s official SDK (@linear/sdk)
• Uses MCP SDK (@modelcontextprotocol/sdk 1.4.0)
• Authentication via API tokens
• Comprehensive error handling
• Rate limiting considerations
• Bun runtime for improved performance
• ESM modules throughout
• Vite build system
• Type-safe operations
• Data cleaning features:
  • Issue mention extraction (ABC-123 format)
  • User mention extraction (@username format)
  • Markdown content cleaning
  • Content optimization for AI context
• Self-assignment support:
  • Automatic current user resolution
  • ‘me’ keyword support in create/update operations
  • Efficient user ID caching
• Advanced search capabilities:
  • Comprehensive filtering with Linear’s API
  • Support for all field comparators
  • Relationship filtering
  • Logical operators (and, or)
  • Relative date filtering
  • Filter by assignee/creator (including self)
  • Support for specific user IDs
  • Project filtering by ID or name
  • Efficient query optimization
• Project management features:
  • Project listing with filtering and pagination
  • Project update creation with health status tracking
  • Project update retrieval with filtering options

Error Handling

The server implements a comprehensive error handling strategy:

• Network error detection and appropriate messaging
• HTTP status code handling
• Detailed error messages with status codes
• Error details logging to console
• Input validation for all parameters
• Label validation and synchronization
• Safe error propagation through MCP protocol
• Rate limit detection and handling
• Authentication error handling
• Invalid query handling
• Team inheritance validation for subissues
• User resolution validation
• Search filter validation

LICENCE

This project is licensed under the MIT License - see the LICENCE file for details.