get_schema
Get the schema of the Bauplan data store.
Query Bauplan's Iceberg tables from the conversation — list tables, read schemas, and run read-only SQL.
LABSCO SUMMARY
The three tools follow the natural order of exploring data — see what tables exist, read the one you care about, then query it — and run_query is limited to SELECT, so an agent can answer questions from the data without any risk of changing it.
A minimalist server for Bauplan, which manages Iceberg tables in S3, giving an assistant three tools to explore and query the data in a configured namespace.
A Bauplan API key, plus the branch and namespace to work in; supports both stdio and SSE transports. Runs through uvx on Python.
One command plus a key — uvx mcp-bauplan, then supply credentials
USE CASES
list_tables shows the tables and get_schema reveals a table's columns, so a query can be written against real structure.
run_query runs a SELECT and returns the rows, read-only by design.
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
get_schemaGet the schema of the Bauplan data store.
list_tablesList all data tables in the Bauplan data store.
run_queryQuery the Bauplan data store using SQL.
BEFORE YOU INSTALL
3 tools when we connected. Queries are SELECT only; pipeline execution, query generation and resources are work in progress, not shipped.
The documented client runs the server on your machine.
CONTEXT COST
Rated by the size of the 3 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
uvx mcp-bauplan{
"mcpServers": {
"mcp-bauplan": {
"command": "<YOUR_PATH_TO_UVX>",
"args": [
"mcp-bauplan"
],
"env": {
"BAUPLAN_API_KEY": "<YOUR_BAUPLAN_API_KEY>",
"BAUPLAN_BRANCH": "<YOUR_BAUPLAN_BRANCH>",
"BAUPLAN_NAMESPACE": "<YOUR_BAUPLAN_NAMESPACE>",
"BAUPLAN_TIMEOUT": "query-timeout-secs"
}
}
}
}