Enables read-only SQL querying and exploration of data files (CSV, Parquet, JSON, Excel, etc.) via DuckDB, supporting local paths, globs, URLs, and S3 buckets.
A minimal MCP server that provides a persistent DuckDB SQL engine to AI assistants, enabling natural-language querying of CSVs, Parquet, and cloud data with 12 tools and optional read-only mode.
Run declarative PySpark transformations on DuckDB datasets for heavy aggregations, using a fixed set of validated operations like filter, select, groupBy, and orderBy.
List all configured database connections to discover available options and select one for use. Returns connection names, types, and details without exposing passwords.
Execute parameterized read-only SQL queries against cataloged fitness-data views to retrieve paginated results for a chosen athlete, with safeguards that reject mutations, filesystem access, and system tables.
Join two spatial datasets using intersects, within, or contains predicates, with automatic engine selection and warnings for empty or non-overlapping results.
Execute SQL queries against Rapid7 vulnerability, asset, policy, and remediation data to filter, aggregate, and join tables for targeted security analysis.
Verify the OmniData server is running, can reach DuckDB, and reports guardrail configuration (row limits, timeout, DB path). Use first to confirm the MCP connection is alive.
Generate summary statistics for each column in a dataset: data type, null percentage, distinct count, min/max, and numeric quartiles. Uses DuckDB to profile large tables efficiently.
Execute data quality validation rules from a YAML file against DuckDB, BigQuery, Athena, Databricks, or Postgres, returning a JSON report with optional LLM-driven root cause analysis.