How do I handle file uploads for Replicate models?

Use the REPLICATE_CREATE_FILE tool to upload base64-encoded files. The tool returns a file ID or URL that can be passed as an input to any model requiring binary data.

How do I find the correct input parameters for a specific model?

Run the REPLICATE_MODELS_GET tool with the model owner and name. This returns the openapi_schema which details every required and optional input key.

Can I run long-running predictions without blocking?

Yes. For long-running jobs, you can omit the 'wait_for' parameter and either use a webhook URL for completion notifications or poll the REPLICATE_PREDICTIONS_LIST tool.

Do I need an API key to use this skill?

Yes, you must have a Replicate account and provide your API token when connecting the Composio MCP server (rube.app/mcp) to your Claude client.

Replicate AI Automation

Name: Replicate AI Automation
Author: Tako3832

byTako3832

0•

Data Science & ML

Automates AI model inference and management on Replicate, enabling seamless integration of image, text, and audio generation into development workflows.

The Replicate Automation skill empowers developers to programmatically interact with thousands of open-source and proprietary AI models directly within their Claude environment via the Composio MCP. By streamlining tasks like model schema inspection, file uploads for multi-modal inputs, and asynchronous prediction tracking, it eliminates the manual overhead of switching between documentation and code. This skill is particularly valuable for developers building AI-native applications who need to test model outputs, validate input schemas, or manage complex inference pipelines without leaving their terminal.

Key Features

01Support for both synchronous waiting and webhook-based asynchronous workflows

02Manage prediction history and track model versioning

03Upload local files as base64-encoded inputs for multi-modal models

04Execute predictions on public models including Llama, Flux, and SDXL

05Inspect model schemas and documentation directly via CLI

060 GitHub stars

Use Cases

01Generating UI assets and placeholder images during frontend development

02Prototyping AI features by testing different model versions and parameters in real-time

03Automating media processing tasks like audio transcription and video generation

Key Features

01Support for both synchronous waiting and webhook-based asynchronous workflows

02Manage prediction history and track model versioning

03Upload local files as base64-encoded inputs for multi-modal models

04Execute predictions on public models including Llama, Flux, and SDXL

05Inspect model schemas and documentation directly via CLI

060 GitHub stars

Use Cases

01Generating UI assets and placeholder images during frontend development

02Prototyping AI features by testing different model versions and parameters in real-time

03Automating media processing tasks like audio transcription and video generation