ritual-dapp-multimodal

Generate images, audio, and video via Ritual two-phase async precompiles.

62|56|Updated Apr 24, 2026
One-click install
npx skills add https://github.com/ritual-foundation/ritual-dapp-skills --skill ritual-dapp-multimodal
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ritual-dapp-multimodal
Source: https://github.com/ritual-foundation/ritual-dapp-skills/tree/main/skills/ritual-dapp-multimodal
Command: npx skills add https://github.com/ritual-foundation/ritual-dapp-skills --skill ritual-dapp-multimodal

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

On-chain multimodal generation capabilities for Ritual dApps, enabling developers to generate images, audio, and video directly from smart-contract-driven requests and publish results to external storage.

Core Features & Use Cases

  • Async, two-phase multimodal generation via precompiles for Image, Audio, and Video with task IDs and callbacks.
  • Storage integration using StorageRef and ECIES-encrypted credentials to upload outputs to GCS/HuggingFace/Pinata, with on-chain verification and secure delegation patterns.
  • Use case: Build on-chain dApps that render generated media in wallets or dashboards by submitting prompts and consuming resulting URIs.

Quick Start

Submit a multimodal generation request to the Ritual precompiles by specifying a prompt, model, desired dimensions or duration, storageRef, and encrypted secrets.

Frequently Asked Questions about ritual-dapp-multimodal

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

FAQPage Schema
How do I generate and store images on-chain for Ritual dApps?

On-chain multimodal generation for Ritual dApps is enabled through async, two-phase precompiles that submit image, audio, or video requests and upload outputs to external storage using StorageRef.

How does async multimodal generation work with two-phase precompiles?

Two-phase precompiles handle multimodal generation by first accepting a prompt and parameters to create a task ID, then delivering resulting media URIs through phase-2 callbacks upon completion.

Can I use ECIES-encrypted credentials to store generated media on GCS or Pinata?

Yes, multimodal generation outputs can be uploaded to GCS, HuggingFace, or Pinata using StorageRef and ECIES-encrypted credentials, enforcing secure delegation and on-chain verification.

What security patterns are enforced for on-chain multimodal generation?

On-chain multimodal generation enforces security patterns like SecretsAccessControl and TEEServiceRegistry lookup to securely manage encrypted credentials and verify authorized service access.

Do I need predefined models for audio and video generation on-chain?

Yes, on-chain multimodal generation requires specifying predefined Image, Audio, or Video models when submitting prompts and desired dimensions or durations through the precompile requests.