multimodal-agents-course
Kubrick AI is an MCP Multimodal Agent designed for video processing tasks.
This course teaches developers how to build an MCP server specifically for video processing using tools like Pixeltable and FastMCP. Participants will learn to design a custom, Groq-powered agent that connects to their MCP server with its own client, enabling advanced functionalities. Additionally, the course covers integration with Opik for enhanced observability and prompt versioning, allowing for a more robust AI system development experience. Throughout the course, learners will gain insights into creating complex MCP servers, implementing custom clients, and utilizing stateful agents for multimodal data processing. By the end, participants will be equipped with the knowledge to build production-ready AI systems that leverage the power of video processing and multimodal capabilities.
Compatible with
Install
Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"multimodal-agents-course": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-multimodal-agents-course"
]
}
}
}Config File Location
Mac: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
Linux: ~/.config/claude/claude_desktop_config.json
Some servers require additional setup - check the GitHub README for specific instructions.
Permissions