yandex-automation

Yandex Automation - Automating Yandex Tasks via Rube MCP

Skill Overview

The Yandex Automation skill enables you to automate Yandex services through Rube MCP and the Composio toolkit, without requiring API keys, supporting tool discovery, connection management, and task execution.

Use Cases

1. Integrating Yandex Features in the Claude Environment
When you need to call various Yandex services directly within a Claude conversation, this skill provides a standardized MCP interface that supports the automation of multiple Yandex services, including mail, disk, and search. It is particularly well suited for building intelligent assistants and workflow automation scenarios.

2. Developing Yandex Integrations Without API Key Authentication
For developers who want to quickly integrate Yandex features without dealing with complex API key management, Rube MCP provides a plug-and-play solution. Simply add the https://rube.app/mcp endpoint to your MCP configuration, then use Composio for unified authentication.

3. Yandex Batch Tasks and Workflow Automation
When you need to handle large volumes of Yandex operations, such as batch email processing or file management, the RUBE_REMOTE_WORKBENCH provided by this skill supports batch execution. Combined with automatic tool discovery and dynamic schema retrieval, it can adapt to changes in the Yandex toolkit.

Core Features

1. Intelligent Tool Discovery and Schema Retrieval
RUBE_SEARCH_TOOLS dynamically queries the available Yandex tools, input parameters, and execution plans, ensuring that the latest tool schemas are always used. For scenarios requiring complete schema definitions, RUBE_GET_TOOL_SCHEMAS can retrieve detailed schema references, avoiding the obsolescence caused by hardcoding.

2. Connection Status Management and Validation
RUBE_MANAGE_CONNECTIONS provides real-time checks of Yandex connection status and supports monitoring states such as ACTIVE, PENDING, and EXPIRED. It automatically verifies connection availability before task execution, preventing operation failures caused by invalid authentication.

3. Flexible Tool Execution Engine
RUBE_MULTI_EXECUTE_TOOL supports individual and batch calls to Yandex tools, automatically handling memory session state and session reuse. Its built-in pagination mechanism ensures that large amounts of data are retrieved completely, making it suitable for scenarios ranging from one-time queries to complex workflows.

Frequently Asked Questions

What prerequisites are required for the Yandex Automation skill?

Three basic requirements are needed: the Rube MCP server must be connected, which can be verified through the RUBE_SEARCH_TOOLS response; the Yandex service must be authorized through RUBE_MANAGE_CONNECTIONS, with its status set to ACTIVE; and tool discovery must be performed before each execution to retrieve the current schema. The first-time setup requires adding the Rube endpoint to the MCP client configuration and authorizing a Yandex account.

How can I check whether the Yandex connection is working properly?

Call RUBE_MANAGE_CONNECTIONS and specify yandex as the toolkit. The connection_status field in the response shows the current status. ACTIVE indicates that the connection is available, PENDING indicates that authorization is still in progress, and EXPIRED indicates that reauthorization is required. It is recommended to perform this check before starting each workflow.

What is the difference between RUBE_SEARCH_TOOLS and RUBE_GET_TOOL_SCHEMAS?

RUBE_SEARCH_TOOLS is used for tool discovery. It returns a list of available tools, basic parameters, and recommended usage, making it suitable for quickly finding which tool to call. RUBE_GET_TOOL_SCHEMAS is used to retrieve complete technical schema definitions, including all field types, validation rules, and nested structures. It should be used when precise parameter construction or debugging is required. Because the Yandex toolkit may be updated, it is recommended to always search first and then execute.