AgentCC
Provides image captioning, object detection, and visual question answering via the Moondream vision model through MCP.