Replicate Automation

Automate Replicate AI model operations including predictions, file uploads, and model schema inspection.

74.2k|8.5k|Updated Oct 17, 2025
One-click install
npx skills add https://github.com/ComposioHQ/awesome-claude-skills --skill replicate-automation
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Replicate Automation
Source: https://github.com/ComposioHQ/awesome-claude-skills/tree/main/composio-skills/replicate-automation
Command: npx skills add https://github.com/ComposioHQ/awesome-claude-skills --skill replicate-automation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill removes the manual overhead of interacting with the Replicate API, allowing you to trigger complex AI model workflows directly from your agentic environment.

Core Features & Use Cases

  • Model Inference: Run predictions on public AI models for image, video, and text generation.
  • Workflow Management: Inspect model schemas, manage file uploads, and track prediction history programmatically.
  • Use Case: Automatically generate images using a specific model version, upload the resulting file to a storage bucket, and log the prediction metadata for audit purposes.

Quick Start

Use the Replicate Automation skill to run a prediction on the flux-1.1-pro model with the prompt 'A sunset over mountains'.

Frequently Asked Questions about Replicate Automation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a prediction on a Replicate model?

Use the REPLICATE_MODELS_PREDICTIONS_CREATE tool by providing the model owner, model name, and the required input object. Ensure you check the model schema first using REPLICATE_MODELS_GET to identify correct input keys.

How can I inspect a model's input requirements?

Call the REPLICATE_MODELS_GET tool with the model owner and name to retrieve the openapi_schema. This schema details the required input keys and data types for that specific model.

Does this skill support long-running model predictions?

Yes, for long-running jobs, omit the wait_for parameter and use webhooks or poll the REPLICATE_PREDICTIONS_LIST tool to track completion status asynchronously.

How do I upload files for model input?

Use the REPLICATE_CREATE_FILE tool to upload binary files. You must provide the content as a base64-encoded string along with the filename and content_type.

What are the limitations of the wait_for parameter?

The wait_for parameter supports a synchronous wait time of 1 to 60 seconds only. Jobs exceeding this duration require asynchronous handling via webhooks or polling.