Google BigQuery

Runs an arbitrary SQL query against a BigQuery dataset and returns the result rows, enabling agents to pull exactly the data they need on demand.

Appends a single new record to a BigQuery table, making it the go-to action for logging one event or entity at a time.

Sends a streaming insert to a BigQuery table for low-latency, high-frequency data ingestion without the overhead of a full load job.

Batch-inserts multiple records into a BigQuery table in one call, dramatically reducing API overhead when loading dozens or hundreds of rows at once.

Removes one or more rows from a BigQuery table that match specified criteria, supporting data hygiene, compliance deletions, and stale-record cleanup.

Loads data from an external source (CSV, JSON, Avro, Parquet, or Cloud Storage file) into a BigQuery table via a managed load job.

Submits a SQL query as an asynchronous BigQuery job and returns a job ID for subsequent status polling, ideal for long-running analytical queries.

Modifies existing records in a BigQuery table that match a filter condition, enabling agents to keep warehoused data in sync with upstream changes.

Retrieves the first row from a BigQuery table that matches the specified filter, enabling quick lookups before branching agent logic.

Fetches the result rows produced by a completed BigQuery job using its job ID, closing the loop on queries submitted asynchronously.

Looks up a row by a unique key and returns it if found, or inserts a new record and returns the created row—eliminating duplicate-check boilerplate from agent workflows.

Returns all tables within a specified BigQuery dataset, giving agents a live catalogue of available data assets before deciding which table to query or update.

Retrieves the schema, metadata, and statistics for a single BigQuery table, letting agents inspect column definitions and row counts before building queries.

Provisions a new table inside a BigQuery dataset with a defined schema, enabling agents to scaffold data storage dynamically as part of pipeline setup or onboarding flows.

Creates a new BigQuery dataset (namespace) within a GCP project, giving agents the ability to provision logical data groupings as part of automated infrastructure or onboarding flows.

Permanently removes a BigQuery table and all its data, enabling agents to clean up temporary tables, enforce retention policies, and decommission deprecated data assets.

Fires when a BigQuery query job finishes and automatically returns the result rows inline, so downstream steps can process data immediately without a separate fetch call.

Fires each time a new row is appended to a monitored BigQuery table, letting agents react to fresh data the moment it lands in the warehouse.

Fires whenever any BigQuery job (query, load, copy, or export) finishes successfully, enabling agents to chain downstream logic onto long-running data operations.

Fires when an existing row in a BigQuery table is modified, letting agents detect changes and propagate updates to downstream systems in near real time.

Fires when a new table is created in a monitored BigQuery dataset, letting agents react immediately to schema provisioning events—useful for triggering validation, notifications, or downstream pipeline setup.

Fires when a previously submitted BigQuery query job completes and its results are available for retrieval, enabling event-driven processing of asynchronous analytical workloads.