MCP tools#

sddb-mcp serves the catalog to agents over stdio (Model Context Protocol). It needs the optional extra:

pip install "spatialdata-db[mcp] @ git+https://github.com/theislab/spatialdata-db-client@v0.1.3"
sddb-mcp

The query tool caps its output at MAX_ROWS = 200 records; genes is not capped. download writes only to a server-controlled directory: $SDDB_MCP_DOWNLOAD_DIR, or a cache subdirectory when unset.

Tool	Description	Parameters
`query`	Filter datasets by facets, ranges (min_obs/min_features), gene (expressing) and free-text search.	`organism`, `tissue`, `disease`, `technology`, `assay`, `tier`, `validation` (default `pass`), `expressing`, `min_fraction`, `search`, `min_obs`, `min_features`
`describe`	Catalog row and element shapes for one dataset uid.	`uid`
`genes`	Datasets expressing one or more genes (symbols or Ensembl ids; mode all=AND, any=union), optionally filtered by organism/tissue/disease/assay. Bronze datasets are included by default.	`symbols`, `organism`, `tissue`, `disease`, `assay`, `mode` (default `all`), `include_bronze` (default true)
`features`	Datasets measuring one or more features of any modality (genes, protein targets, metabolites; labels or ids; mode all=AND, any=union), optionally filtered by modality/feature_type and organism/tissue/disease/assay. Each row lists the typed feature hits.	`features`, `modality`, `feature_type`, `organism`, `tissue`, `disease`, `assay`, `mode` (default `all`), `include_bronze` (default true)
`facets`	List facet columns, or the distinct values of a facet field.	`field` (optional)
`cite`	BibTeX for every study of a filtered cohort. Studies without a published citation get a fallback entry that cites the dataset page; `n_fallback` counts them.	`organism`, `tissue`, `disease`, `technology`, `assay`, `tier`, `search`, `expressing`, `min_fraction`, `validation`
`download`	Plan (sizes) or fetch a filtered cohort to the server download dir.	same filters as `cite`, plus `download` (default false: plan only)