execute_query
Execute a SQL query on the Impala database and return results as JSON.
LABSCO SUMMARY
It hands SQL straight to Impala — there is no query builder in between.
get_schema returns table names and nothing else, so column layouts have to come out of the queries themselves. execute_query takes only a query string, which means paging, limiting and aggregation are things you write into the SQL rather than pass as arguments. Two tools and JSON results make it easy to chain, and equally easy to pull back more rows than you wanted.
What it is
A SQL bridge to an Impala database over Iceberg tables, with one tool for table discovery and one for query execution.
What you get
- Table discovery through get_schema, which returns the table names in the current Impala database.
- Arbitrary SQL execution through execute_query, with results returned as JSON.
Requirements
An Impala database the server can reach, plus uv and a Python for it to run under — that is what UV_PYTHON selects.
Setup effort
Add the configuration to your client
USE CASES
01
Getting oriented in a warehouse nobody documented
The table names are the starting point when there is no data dictionary to read, and they are what any subsequent query has to be written against.
02
Settling a number mid-conversation without opening a SQL client
A query goes straight to Impala and the rows come back as JSON, which keeps the answer in the same place as the question that prompted it.
2 tools
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
get_schema
Retrieve the list of table names in the current Impala database.
BEFORE YOU INSTALL
What to know first
Tools you get
2 tools when we connected.
Requires
- IMPALA_PASSWORD in your environment
- IMPALA_USER in your environment
Where it runs
The documented client runs the server on your machine.
CONTEXT COST
Light Standing cost
Rated by the size of the 2 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Connect your client
MCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"iceberg-mcp-server": {
"command": "uvx",
"args": [
"--from",
"git+https://github.com/cloudera/iceberg-mcp-server@main",
"run-server"
],
"env": {
"IMPALA_HOST": "<YOUR_IMPALA_HOST>",
"IMPALA_PORT": "443",
"IMPALA_USER": "<YOUR_IMPALA_USER>",
"IMPALA_PASSWORD": "<YOUR_IMPALA_PASSWORD>",
"IMPALA_DATABASE": "default"
}
}
}
}
