Paradym Automation - Automate Paradym Tasks via Rube MCP

Skill Overview

Paradym Automation is a skill that automates Paradym operations through Composio’s Rube MCP service. It provides capabilities for tool searching, connection management, and batch execution.

Use Cases

1. Paradym Workflow Automation: Programmatically execute Paradym tasks, such as batch operations, scheduled jobs, or complex workflow orchestration.
2. MCP Integration Development: Building MCP-based applications that need Paradym functionality integrated—without manually managing API keys.
3. Dynamic Tool Discovery: Paradym tool schemas may change in a project, requiring dynamic retrieval of tool definitions at runtime instead of hard-coding them.

Core Features

1. Intelligent Tool Search: Use RUBE_SEARCH_TOOLS to dynamically discover available Paradym tools, retrieve real-time schemas, and execution plans.
2. Connection Status Management: Use RUBE_MANAGE_CONNECTIONS to verify the Paradym connection status, supporting authorization links and activation checks.
3. Batch Task Execution: Use RUBE_MULTI_EXECUTE_TOOL and RUBE_REMOTE_WORKBENCH to perform bulk tool calls and remote workbench operations.

FAQs

What prerequisites does Paradym Automation require?

This skill requires three prerequisites: (1) the Rube MCP server is configured (added in the client as https://rube.app/mcp), (2) the Paradym connection is activated (verified via RUBE_MANAGE_CONNECTIONS), and (3) before each execution, call RUBE_SEARCH_TOOLS to fetch the latest tool schema. No API key is required—Rube MCP provides keyless authentication via Composio.

Why must RUBE_SEARCH_TOOLS be called before executing tools?

The Paradym tool schema (field names, types, required parameters) may change at any time. Hard-coding tool definitions can cause calls to fail. RUBE_SEARCH_TOOLS returns the currently available tool list, input schemas, recommended execution plans, and known pitfalls—ensuring your calls remain compatible with the latest API.

How do I handle Paradym workflows that require multiple steps?

Use the session ID reuse mechanism: when searching for tools in the first step, generate a session ID (generate_id: true), and reuse the same session ID for subsequent connection checks and tool execution. This ensures Rube MCP correctly associates the context. For batch operations, use the run_composio_tool() method of RUBE_REMOTE_WORKBENCH, or fully retrieve all data by handling the pagination token in the response.