# Dexto Agent Configuration for Sora Video Generation MCP Server # Optional greeting shown at chat start (UI can consume this) greeting: "🎬 Hello! I'm your Sora Video Agent. Let's create some amazing videos together!" systemPrompt: | You are an AI assistant specialized in video generation using OpenAI's Sora technology. You have access to a comprehensive set of tools for creating, managing, and manipulating AI-generated videos including: - **Video Generation**: Create videos from text prompts with custom duration, resolution, and style - **Reference-Based Creation**: Use images or videos as reference for more precise generation - **Video Management**: Monitor generation progress, list all videos, and organize your creations - **Video Remixing**: Create variations and extensions of existing videos with new prompts - **File Management**: Automatically download and organize generated videos - **Quality Control**: Delete unwanted videos and manage storage efficiently When working with video generation: 1. Always provide clear, detailed prompts for the best results 2. Consider the target audience and use case when choosing duration and resolution 3. Monitor video generation progress and provide status updates 4. Suggest creative variations and remixes when appropriate 5. Help users organize and manage their video library 6. Provide guidance on optimal video settings for different use cases **Supported Video Specifications:** - **Durations**: 4s, 8s, 16s, 32s - **Resolutions**: 720x1280 (9:16), 1280x720 (16:9), 1024x1024 (1:1), 1024x1808 (9:16 HD), 1808x1024 (16:9 HD) - **Reference Formats**: JPG, PNG, WebP, MP4, MOV, AVI, WebM **Best Practices:** - Use descriptive, specific prompts for better results - Consider the aspect ratio for your intended platform (vertical for social media, horizontal for YouTube) - Start with shorter durations for testing, then create longer versions - Use reference images/videos for consistent style or character continuity - Monitor generation progress as it typically takes 1-3 minutes - Save completed videos promptly to avoid losing access mcpServers: sora_video: type: stdio command: npx args: - -y - "@truffle-ai/sora-video-server" connectionMode: strict env: OPENAI_API_KEY: $OPENAI_API_KEY toolConfirmation: mode: "auto-approve" allowedToolsStorage: "memory" llm: provider: openai model: gpt-5-mini apiKey: $OPENAI_API_KEY internalResources: enabled: true resources: - type: blob storage: cache: type: in-memory database: type: sqlite blob: type: local # CLI provides storePath automatically maxBlobSize: 209715200 # 200MB per blob (for HD videos up to 32 seconds) maxTotalSize: 2147483648 # 2GB total storage cleanupAfterDays: 30 # Prompts - video generation examples shown as clickable buttons in WebUI prompts: - type: inline id: create-video title: "🎬 Create Video" description: "Generate video from text prompt" prompt: "Create an 8-second video of a cat playing piano in a cozy living room with warm lighting." category: generation priority: 10 showInStarters: true - type: inline id: nature-video title: "🌄 Nature Video" description: "Generate scenic landscape video" prompt: "Generate an 8-second video in 16:9 format showing a breathtaking sunset over mountains with clouds moving across the sky." category: nature priority: 9 showInStarters: true - type: inline id: social-video title: "📱 Social Media Video" description: "Create vertical video for social platforms" prompt: "Create a 4-second vertical video (9:16) showing a product reveal with dramatic lighting." category: social-media priority: 8 showInStarters: true - type: inline id: futuristic-video title: "🚀 Futuristic Scene" description: "Generate sci-fi themed video" prompt: "Create a 16-second video showing a futuristic cityscape with flying cars and neon lights at night." category: sci-fi priority: 7 showInStarters: true