textcortex-automation

Textcortex Automation - Automating Text Generation via Rube MCP

Skill Overview

Automate Textcortex text generation tasks through Composio’s Textcortex toolkit and the Rube MCP server, supporting tool discovery, connection management, and batch operation orchestration.

Applicable Scenarios

AI Writing Workflow Automation

For scenarios requiring the batch generation of copy, summaries, or creative content, tool orchestration enables continuous text generation tasks without manual intervention.

Enterprise Text Processing Integration

Integrate Textcortex capabilities into existing applications and provide unified text generation services through the MCP standard interface, without directly handling API authentication or rate limits.

Multi-Tool Collaborative Automation

Combine other MCP tools, such as search and data analysis, to build complete intelligent workflows, with Textcortex embedded as the text generation component within the overall process.

Core Features

Dynamic Tool Discovery and Schema Validation

Use RUBE_SEARCH_TOOLS to retrieve the latest list of Textcortex tools and input parameter structures in real time, avoiding the risk of schema changes caused by hardcoding. Supports searching for recommended tools and known pitfalls based on use cases.

Connection Status Management and Authentication

Use RUBE_MANAGE_CONNECTIONS to centrally manage Textcortex authentication status and automatically detect connection activity. When authentication expires, users are guided through the authorization process to ensure stable workflow execution.

Session-Based Batch Execution

Use RUBE_MULTI_EXECUTE_TOOL to execute multiple Textcortex operations within the same session, supporting context retention and cross-tool parameter passing. This is suitable for complex text generation task chains.

Frequently Asked Questions

How do I connect Textcortex through Rube MCP?

First, ensure that the Rube MCP server (https://rube.app/mcp) has been added to the client configuration. Call RUBE_MANAGE_CONNECTIONS and specify the toolkit as "textcortex". If the connection is inactive, the system will return an authentication link. Once authorization is completed, the toolkit can be used. No manual API key configuration is required.

Can the schemas of Textcortex tools change?

Yes. Therefore, you must always call RUBE_SEARCH_TOOLS first to obtain the latest tool definitions and parameter structures. This skill requires tool discovery before executing any operation. The returned results include a recommended execution plan and known considerations to help prevent call failures caused by schema changes.

How can I perform batch Textcortex operations?

Use RUBE_MULTI_EXECUTE_TOOL to orchestrate multiple tool calls in a single request, reusing session state through session_id. For more complex batch tasks, use RUBE_REMOTE_WORKBENCH to execute composio_tool() scripts. When making calls, always include the memory parameter, even if it is an empty object, as required by the interface specification.

What is the difference between RUBE_SEARCH_TOOLS and RUBE_GET_TOOL_SCHEMAS?

RUBE_SEARCH_TOOLS is used to discover available tools. It supports fuzzy searches by use case and returns a list of tools, recommended usage, and known pitfalls. RUBE_GET_TOOL_SCHEMAS retrieves the detailed schema definitions of specific tools and only needs to be called when the search results contain a schemaRef. Both are necessary steps in the workflow.

How should the session ID be reused in a workflow?

Set generate_id to true in the initial RUBE_SEARCH_TOOLS call to create a session. All subsequent RUBE_MANAGE_CONNECTIONS and RUBE_MULTI_EXECUTE_TOOL calls should use the returned session_id. Operations within the same session can share context and authentication status. A new session ID should be generated for each new workflow.

How can I verify whether the Textcortex connection is available?

Call RUBE_MANAGE_CONNECTIONS before executing any tools and check whether the connection status is ACTIVE. If the status is inactive, complete reauthorization using the returned auth_link. It is recommended to check the connection at the beginning of every workflow to avoid authentication failures midway through execution.