Replicate AI Handler Skill This skill provides a production-ready pattern for integrating Replicate AI models. It handles long-running predictions using Inngest background jobs, stores results in S3, manages user credits, and updates database state. Architecture 1. Trigger : User requests a generation via API (e.g., ). 2. Validation : Check/deduct user credits. 3. State : Create a database record with . 4. Queue : Trigger an Inngest function to handle the Replicate API call. 5. Processing : - Call Replicate API. - Wait for completion (polling or webhook). - Download result and upload to S3 (s…