wodeapp-ai_skill

Overview

This skill exposes the WodeApp AI Engine: a unified, token-efficient platform that unlocks text, image, video, TTS, structured JSON, visual workflows, and no-code page building with zero-deploy publishing. It gives one API key and a single credit pool for access to 343+ models and MCP-native auto-discovery. The pack is designed for rapid integration, low operational overhead, and multi-modal output pipelines.

How this skill works

The skill routes requests to WodeApp runtime endpoints (HTTP or MCP SSE) using an X-API-Key header and a unified credit pool. It offers dedicated endpoints and MCP tools for text, image, video, TTS, JSON schema responses, workflow execution, and page publishing. Visual workflows and no-code page builder allow drag-and-drop pipelines and one-click publish to a managed domain with SSL.

When to use it

• Prototype multi-modal AI features quickly without managing models or infra
• Generate structured JSON outputs that follow your schema for downstream processing
• Create images or dynamic videos (Seedream / Seedance / Kling) from prompts or assets
• Build and publish interactive pages and AI workflows with zero-deploy
• Consolidate billing and quotas across many models into one credit pool
• Integrate into MCP-enabled apps for auto-discovered tools and low-latency streaming

Best practices

• Store the API key in environment variables and never hardcode credentials
• Use project-scoped or scoped keys with quotas for production and team separation
• Prefer MCP (SSE) for auto-discovery of tools and smoother streaming UX
• Define explicit JSON schemas in structured endpoints to ensure deterministic outputs
• Monitor per-project rate and credit usage and set alerts for HTTP 402 billing failures
• Revoke suspected compromised keys immediately from the WodeApp console

Example use cases

• Automated content pipeline: form → AI copy → review → publish an interactive page
• Multi-modal campaign assets: generate brand taglines, hero images, and short videos in one flow
• Voice-enabled assistants: synthesize multi-language TTS with selectable voices for apps
• Structured prompt generation: produce validated JSON prompts for downstream model orchestration
• Rapid prototyping: drag-and-drop workflows to test model combos and publish demos without ops