Skip to main content
MCP (Model Context Protocol) servers extend your agents with additional tools, integrations, and capabilities. Tembo supports both built-in MCP servers and custom configurations.

​
What is MCP?

The Model Context Protocol is an open standard that allows AI agents to connect to external tools, APIs, and data sources. MCP servers provide tools that your agents can use to interact with external systems, from browser automation to project management integrations. Learn more at modelcontextprotocol.io.

​
Built-in MCP Servers

Tembo provides several MCP servers out of the box:
ServerTypeDescription
TembostdioCore Tembo tools (sub-issues, Slack, tasks)
PlaywrightstdioBrowser automation and web testing
Context7stdioContext management and search
KernelhttpCloud browser infrastructure for web agents

​
Tembo MCP

The Tembo MCP server provides tools for interacting with Tembo’s features:
  • create_task - Create new tasks with optional repository and branch specifications
  • list_tasks - List and paginate tasks for your organization with filtering options
  • search_tasks - Search tasks by title or description content
  • list_repositories - Display enabled code repositories for your organization
  • get_current_user - Get authenticated user information
  • createSubIssue - Create sub-issues from tasks (available when used within Tembo)
  • sendSlackMessage - Send messages to Slack channels (requires Slack integration, available when used within Tembo)
The Tembo MCP server is available in two ways:
  1. Built-in - Automatically available for all Tembo agents
  2. Public npm package - Can be used with other MCP clients like Claude Desktop

​
Using Tembo MCP with Claude Desktop

The Tembo MCP server is published as a public npm package, allowing you to use Tembo’s task management features directly in Claude Desktop or other MCP-compatible applications. Installation You can use the Tembo MCP server without installation via npx: 
npx -y @tembo-io/mcp
Or install it globally: 
npm install -g @tembo-io/mcp
Configuration
  1. Generate a Tembo API key from your API Keys settings
  2. Add the following configuration to your MCP client settings file (for Claude Desktop, this is claude_desktop_config.json): 
{
  "mcpServers": {
    "tembo": {
      "command": "npx",
      "args": ["-y", "@tembo-io/mcp"],
      "env": {
        "TEMBO_API_KEY": "your-api-key-here"
      }
    }
  }
}
Replace your-api-key-here with your actual API key from Step 1.
  1. Restart your MCP client to load the Tembo MCP server
Once configured, you can ask Claude to create tasks, search for existing tasks, list repositories, and more - all directly from Claude Desktop.

​
Playwright MCP

Enables agents to automate browser interactions:
  • Navigate web pages
  • Fill forms and click elements
  • Take screenshots
  • Run automated tests

​
Context7 MCP

Provides enhanced context management capabilities for your agents.

​
Kernel MCP

The Kernel MCP server provides cloud-based browser infrastructure for AI agents and automation tasks. It offers serverless browsers with anti-bot detection, session persistence, and autoscaling capabilities. Key use cases:
  • Web scraping and data extraction - Extract data from websites that require JavaScript rendering or authentication
  • Automated testing - Run end-to-end browser tests in isolated cloud environments
  • Form filling and submissions - Automate complex web workflows that require user interactions
  • Screenshot capture - Generate screenshots of web pages for documentation or monitoring
  • Browser automation at scale - Run hundreds of concurrent browser sessions without infrastructure management
Available tools:
  • create_browser / delete_browser - Launch and terminate browser sessions
  • execute_playwright_code - Run Playwright/TypeScript code with automatic video replay
  • take_screenshot - Capture screenshots with optional region selection
  • setup_profile / list_profiles - Manage browser profiles for session persistence
  • list_apps / invoke_action - Interact with Kernel apps and deployments
  • search_docs - Search Kernel documentation for up-to-date information via MCP query
Configuration: To use Kernel MCP, add your Kernel API key in Settings > Agents under the “Kernel API Key” field. Get your API key from onkernel.com.

​
Integration MCP Servers

When you connect integrations to Tembo, additional MCP servers become available automatically:
IntegrationServer TypeTools Provided
GitHubstdioRepository management, issues, pull requests
LinearhttpIssue tracking, project management
NotionhttpPages, databases, blocks, workspace structure
These servers are automatically configured with the appropriate authentication when you connect the integration.

​
Custom MCP Servers

You can add custom MCP servers in your organization settings. Tembo supports three types of MCP servers:

​
Server Types

stdio - Local process-based servers 
{
  "command": "node",
  "args": ["path/to/server.js"],
  "env": {
    "API_KEY": "your-key"
  }
}
http - HTTP-based servers 
{
  "url": "https://example.com/mcp",
  "headers": {
    "Authorization": "Bearer token"
  }
}
sse - Server-Sent Events servers 
{
  "url": "https://example.com/mcp/sse",
  "headers": {
    "Authorization": "Bearer token"
  }
}

​
Configuration Format

Custom MCP servers can be configured using either object or array format:

​
Object Format

{
  "mcpServers": {
    "my-server": {
      "command": "node",
      "args": ["server.js"],
      "env": {
        "API_KEY": "key"
      }
    },
    "api-server": {
      "url": "https://api.example.com/mcp",
      "headers": {
        "Authorization": "Bearer token"
      }
    }
  }
}

​
Array Format

{
  "mcpServers": [
    {
      "name": "my-server",
      "type": "stdio",
      "command": "node",
      "args": ["server.js"],
      "env": {
        "API_KEY": "key"
      }
    },
    {
      "name": "api-server",
      "type": "http",
      "url": "https://api.example.com/mcp"
    }
  ]
}

​
Adding Custom MCP Servers

To add custom MCP servers to your organization:
  1. Navigate to Settings > Agents in your Tembo dashboard
  2. Scroll to the Custom MCP Servers section
  3. Enter your MCP configuration in JSON format
  4. Press Cmd+Enter (Mac) or Ctrl+Enter (Windows/Linux) to save
  5. New tasks will automatically use the configured MCP servers

​
Best Practices

Security: Store sensitive credentials as environment variables rather than hardcoding them in configuration. Testing: Test MCP servers locally before adding them to your organization configuration. Documentation: Document custom MCP servers for your team, including what tools they provide and when to use them. Version Control: Keep MCP server configurations in version control alongside your rule files for team consistency.

​
Troubleshooting

MCP server not available: Ensure the server configuration is valid JSON and all required fields are provided. Authentication errors: Verify credentials and API keys are correct and have the necessary permissions. Connection issues: For HTTP/SSE servers, check that the URL is accessible and the server is running. For additional help, contact support.