📄 中文摘要
VDCook是一个可配置的视频数据构建平台，旨在为研究人员和特定领域团队提供服务。用户可以通过自然语言查询和可调参数（如规模、检索-合成比、质量阈值）发起数据请求。系统自动执行查询优化，同时运行真实视频检索和受控合成模块，最终生成具有完整来源和元数据的领域内数据包，并附带可重复的Notebook。与传统的静态一次性构建数据集不同，VDCook通过基于模型上下文协议（MCP）的自动数据摄取机制，实现了持续更新和领域扩展。
VDCook：自我演化的视频数据操作系统
出处: VDCook:DIY video data cook your MLLMs
发布: 2026年3月9日
📄 English Summary
VDCook:DIY video data cook your MLLMs
VDCook is introduced as a configurable video data construction platform designed for researchers and vertical domain teams. Users initiate data requests through natural language queries and adjustable parameters such as scale, retrieval-synthesis ratio, and quality threshold. The system automatically performs query optimization while concurrently running real video retrieval and controlled synthesis modules. It ultimately generates in-domain data packages with complete provenance and metadata, along with reproducible Notebooks. Unlike traditional static, one-time-built datasets, VDCook enables continuous updates and domain expansion through its automated data ingestion mechanism based on the Model Context Protocol (MCP).