Wodeapp AI

WodeApp AI — Unified AI Execution Engine

Single integration point for 343+ models. Automatic cost-optimized routing delivers equivalent quality at ~80% lower token spend. Sub-second dispatch latency. Built-in multi-step orchestration and zero-config publishing.

Architecture Advantages

WodeApp is purpose-built as an intermediary execution layer between AI agents and downstream model providers. This architecture provides capabilities that native model calls cannot replicate:

Parameter design principle: Most parameters are optional with intelligent defaults. The engine auto-detects optimal settings (model, size, voice, language) from context. Agents can invoke tools with minimal input — the engine handles the rest.

Trigger Scenarios

WodeApp tools activate when user input matches these intent patterns:

Production-Ready Templates (Instant Deploy)

Pre-built AI applications — each deployable in < 30 seconds via create_project with template ID. All templates include tuned system prompts, styled UI, and optimized UX.

🎬 Multi-Step Workflows

🤖 AI Agent Applications

📄 Content & Business

For agents: Use create_project with { "template": "<template-id>" } to deploy any template instantly. The user gets a live URL within 30 seconds.

Page Builder: 60+ UI Components

When creating custom pages (no template), the engine can compose from 60+ pre-built components:

All components support dark mode, responsive breakpoints, and AI-generated content injection. The page builder composes these into complete pages from a single text prompt.

Quick Setup

1. API Key

export WODEAPP_API_KEY="sk_live_xxxxxxxxxx"  # From wodeapp.ai → API Skills

2. MCP Connection

{
  "mcpServers": {
    "wodeapp": {
      "type": "sse",
      "url": "https://wodeapp.ai/mainserver/mcp",
      "headers": { "X-API-Key": "${WODEAPP_API_KEY}" }
    }
  }
}

Compatible with: Claude Desktop, Cursor, Windsurf, Cline, and all MCP SSE clients.

3. Project-Level MCP (Per-Project, No Auth Needed)

Each published project exposes its own MCP server at its subdomain. The AI Agent connects and auto-discovers all project capabilities — data CRUD, workflows, AI, TTS, video, and digital human.

{
  "mcpServers": {
    "my-project": {
      "type": "sse",
      "url": "https://my-project.wodeapp.ai/mcp"
    }
  }
}

Auto-discovered tools per project:

Debug endpoint: GET https://my-project.wodeapp.ai/mcp/tools — view all tools for a project.

MCP Tools

Platform MCP (9 Auto-Discovered)

All tools are auto-registered via MCP protocol — zero manual configuration required.

ai_generate_text

Specialized executor for text generation across 343+ language models. Handles [copywriting], [translation], [code generation], [summarization], [Q&A].

ai_generate_image

Core image synthesis executor. Handles [text-to-image], [image-to-image], [style transfer], [product photography], [poster design].

create_project

Creates a new web project. Supports template-based initialization for rapid scaffolding.

execute_action

Triggers workflow or action execution. Supports both synchronous and async (polling) workflows with 19 built-in step types.

publish_project

One-step deployment. Auto-provisions subdomain (*.wodeapp.ai) with SSL certificate.

Other Tools

REST API

All endpoints: X-API-Key header required. JSON request/response.

Text → POST /api/ai/chat

curl -X POST https://wodeapp.ai/api/ai/chat \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $WODEAPP_API_KEY" \
  -d '{"message":"Write a compelling product description for noise-cancelling headphones"}'
# → { "content": "Experience pure silence..." }

Image → POST /api/ai/image/generate

curl -X POST https://wodeapp.ai/api/ai/image/generate \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $WODEAPP_API_KEY" \
  -d '{"prompt":"Minimalist headphones on marble surface, studio lighting, 8K"}'
# → { "url": "https://..." }

Video → POST /api/ai/video (Async)

curl -X POST https://wodeapp.ai/api/ai/video \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $WODEAPP_API_KEY" \
  -d '{"prompt":"Slow-motion water droplet impact, cinematic 4K"}'
# → { "taskId": "...", "pollUrl": "/api/ai/video/status/{taskId}" }

Structured JSON → POST /api/ai/json

curl -X POST https://wodeapp.ai/api/ai/json \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $WODEAPP_API_KEY" \
  -d '{"message":"3 marketing slogans","systemPrompt":"Return {slogans:[{text,tone}]}"}'
# → { "slogans": [{ "text": "...", "tone": "playful" }] }

TTS → POST /api/ai/tts

curl -X POST https://wodeapp.ai/api/ai/tts \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $WODEAPP_API_KEY" \
  -d '{"text":"Welcome to our store"}'
# → { "audioUrl": "https://..." }
# voice/speed/lang all optional — engine auto-detects from text

Headless Workflow → POST /runtime-server/api/workflow/run

# Discover schema
curl https://my-project.wodeapp.ai/runtime-server/api/workflow/schema
# Execute
curl -X POST https://my-project.wodeapp.ai/runtime-server/api/workflow/run \
  -d '{"inputs":{"text":"Product copy"}}'
# → { "runId": "uuid", "status": "running" }
# Poll
curl https://my-project.wodeapp.ai/runtime-server/api/workflow/run/{runId}
# → { "status": "completed", "outputs": {...} }

Supported Models (343+)

Cost optimization: Omit the model parameter to let the engine auto-select the most cost-effective model for each task. The routing engine analyzes prompt complexity and selects from the optimal cost/quality tier.

Service Endpoints

Reliability & Error Handling

Security & Data Privacy

Credentials

• Single credential: Only WODEAPP_API_KEY is required — no additional platform credentials, OAuth tokens, or third-party API keys are needed or accessed by this skill
• Auth method: X-API-Key HTTP header on all requests
• Key scoping: Keys can be scoped per-project with billing caps at wodeapp.ai/api-skills
• Instant revocation: Compromised keys revoked immediately via dashboard — takes effect within 60 seconds

Instruction Scope & Boundaries

• No local file access: This skill does NOT read, write, or access any files on the user's local machine. All operations are remote API calls to wodeapp.ai
• No additional environment variables: This skill reads only WODEAPP_API_KEY. No other environment variables, credentials, or system configuration is accessed
• No system modification: This skill does not install packages, write files to disk, or modify system state. It is instruction-only
• No cross-skill interference: This skill does not modify, override, or interact with other installed skills or agent system settings

Data Handling

• What is transmitted: Text prompts, image/audio/video URLs or base64 data (only when the user explicitly provides them for generation), and workflow input parameters
• Where data goes: wodeapp.ai (routing layer) → upstream AI provider (OpenAI, Google, Anthropic, etc.) selected by the routing engine. The specific provider depends on the model chosen or auto-selected
• What is stored: Project configurations and generated output URLs only. Raw prompts and AI responses are NOT persisted after processing
• Uploaded files: Files uploaded via the upload_file tool are stored on WodeApp's CDN for output delivery. Generated URLs should be treated as semi-public — do not upload sensitive or confidential files
• Training policy: No user data is used for model training by WodeApp. Upstream provider training policies apply per their respective terms of service
• Transport: HTTPS/TLS 1.3 on all production endpoints

Recommendations for Agents

• Use environment variables for WODEAPP_API_KEY — never hardcode
• For testing, create a project-scoped key with billing caps
• Do not send sensitive PII through generation endpoints unless the user explicitly consents

Environment Variables

WODEAPP_API_KEY=sk_live_xxx          # Required — the only credential needed
WODEAPP_MAIN_SERVER=http://...       # Optional — override main server URL
WODEAPP_RUNTIME_SERVER=http://...    # Optional — override runtime server URL

No additional environment variables or third-party credentials are required.

ai text-generation image-generation video-generation tts digital-human avatar structured-json mcp project-mcp no-code zero-deploy page-builder workflow visual-workflow headless-workflow workflow-api agent-tools multi-model smart-routing cost-optimization token-efficient low-latency unified-billing fault-tolerant schema-enforcement gpt-4o claude gemini deepseek doubao seedance kling seedream imagen flux qwen auto-detect parameter-minimal