Mobus
An MCP connector that lets an AI assistant search, preview and analyze datasets across 20 platforms such as Zenodo, Hugging Face, arXiv and SEC EDGAR from a single conversation, checking licenses and generating citations.
Mobus is a hosted MCP connector for dataset discovery, added to Claude by pasting its URL into a custom connector with no install or API keys required. Once connected, a user can ask in plain language to search for datasets by topic or license, preview rows from a specific dataset, find similar or related datasets, or check whether a dataset can be used commercially.
Its tools cover discovery across all 20 supported platforms in parallel, evaluation through metadata and side-by-side comparison, quality assessment for missing values and duplicates, license and schema compatibility checks, citation generation in APA, BibTeX or Chicago format, and research tools that trace a dataset's originating paper, its lineage of variants and derivatives, and its citation graph. Supported platforms without authentication include data.gov, Zenodo, OpenML, the UCI ML Repository, AWS Open Data, arXiv, Census.gov, SEC EDGAR and Crossref, while Hugging Face, Kaggle and Google Dataset Search need an API token or key, and requests degrade gracefully when a key is missing.
It is written in TypeScript and self-hostable.