
Read-only data patrol for AI agents: find broken partitions, prove fixes, escalate the rest.
Public tool metadata for what this MCP can expose to an agent.
patrol_duckdbPatrol a fact table in a DuckDB file, with your column namesPatrol a fact table in a DuckDB file, with your column names
No parameter schema in public metadata yet.
patrol_filesPatrol exported CSV, TSV or Parquet filesPatrol exported CSV, TSV or Parquet files
No parameter schema in public metadata yet.
patrol_snowflakePatrol a Snowflake table, reading only the last N daysPatrol a Snowflake table, reading only the last N days
No parameter schema in public metadata yet.
patrol_dbtPatrol the table behind a dbt model, using the mapping in its `meta` blockPatrol the table behind a dbt model, using the mapping in its `meta` block
No parameter schema in public metadata yet.
get_incidentReturn the full evidence for one incident from a previous patrolReturn the full evidence for one incident from a previous patrol
No parameter schema in public metadata yet.
run_demoPatrol generated data with planted problems, to see the outputPatrol generated data with planted problems, to see the output
No parameter schema in public metadata yet.
Sponsors
A park ranger walks the trails, fixes what is safe to fix on the spot, radios in anything dangerous, and files a report. DataRanger does the same for your tables. It finds data-quality incidents, tests every candidate fix on a copy against the source's control totals, applies only the fixes it can prove, and keeps wrong fixes under a rate you choose. Everything else goes to a person, with the evidence attached.
dataranger