Google Cloud said on September 30, 2026 that its Data Agent Kit, a free set of Model Context Protocol tools and agent skills, is now generally available. The company said the kit lets coding agents including Antigravity, Claude Code and Codex work directly with Google Cloud data products.
What the kit does
According to the post by product manager Arun Nair and developer advocate Jeff Nelson, Data Agent Kit combines MCP tools that connect to more than 15 Google Data Cloud services with open-source, Google-authored skills that teach agents data practices such as optimizing BigQuery SQL, designing Bigtable row keys and building dbt pipelines.
Google said coding agents are already good at writing SQL, PySpark and pipeline code but lack context about a user's environment, such as which tables exist, how they are partitioned and why a job failed. The MCP tools let an agent inspect schemas, run queries, read job logs and manage resources in a live environment, the company said.
Google said every step runs with the user's own IAM permissions, and that the agent connects as the user or as a service account it impersonates, so row- and column-level security policies apply automatically.
What is new at general availability
Google said GA adds support for BigQuery Graph, Bigtable and Managed Service for Apache Spark access to the open Lakehouse, plus what it described as dozens of quality-of-life improvements.
For graphs, the company said a bigquery-graph-author skill maps tables to nodes and edges, checks proposed relationships against the data and presents a plan for approval before creating anything, while a bigquery-graph-query skill writes the GQL. Bigtable joins Spanner, AlloyDB and Cloud SQL as a fully supported database, with a new bigtable-basics skill that Google said flags hotspots and full table scans before table creation.
On the Lakehouse side, Google said agents can now work with Apache Iceberg tables using serverless Spark on Managed Service for Apache Spark, with session state preserved across statements and Iceberg features including branching, time travel and schema evolution. A federate-lakehouse-catalog skill connects a Lakehouse to AWS Glue and Databricks Unity Catalog, the company said.
Availability, pricing and controls
Google said Data Agent Kit is included at no additional cost and that customers pay standard pricing only for the Google Cloud services their agent uses. The skills are open source on GitHub, so teams can audit, fork or extend them, the company said.
The kit is available as an IDE extension for VS Code, Antigravity IDE, Cursor and other VS Code-compatible editors, as a plugin for Antigravity 2.0, Antigravity CLI, Claude Code and Codex, and pre-installed in Cloud Shell and Cloud Workstations.
For administrators, Google said MCP access can be governed with Identity and Access Management, MCP traffic screened with Model Armor, and agents scoped with VPC Service Controls and Principal Access Boundary policies.
What to do
- To install the IDE extension, Google says to search for "Google Cloud Data Agent Kit" in the Extensions panel of VS Code, Antigravity IDE, Cursor or any VS Code-compatible editor, or install it from the VS Code Marketplace or Open VSX.
- In Antigravity 2.0, Google says to go to Settings > Customizations > Build with Google Plugins, then download Data Agent Kit.
- In Cloud Shell and Cloud Workstations the extension is already installed by default; Google says to open the editor and sign in.
- For the CLI plugin, Google lists these commands from its data-agent-kit-plugin repository: Antigravity CLI, agy plugin install https://github.com/GoogleCloudPlatform/data-agent-kit-plugin; Claude Code, claude plugin install data-agent-kit-starter-pack@claude-plugins-official; Codex CLI, codex plugin marketplace add GoogleCloudPlatform/data-agent-kit-plugin followed by codex plugin add dak@dak-marketplace.
- At setup, Google says you sign in once and select the Google Cloud services you use, after which the kit enables the required APIs, installs matching skills and configures MCP servers without manual setup files.
Key facts and where they come from
- Data Agent Kit reached general availability on September 30, 2026.
Today, Google Cloud Data Agent Kit is generally available.
- GA adds BigQuery Graph, Bigtable and Managed Service for Apache Spark support.
we are adding support for BigQuery Graph, Bigtable, and Managed Service for Apache Spark access to your open Lakehouse
- The MCP tools connect to more than 15 Google Data Cloud services.
Connections to more than 15 Google Data Cloud services, so your agent can inspect schemas, run queries, read job logs, and manage resources in your live environment.
- The kit is free, with users paying only standard service pricing.
Data Agent Kit is included at no additional cost; you pay standard pricing only for the Google Cloud services your agent uses.
- Agent actions run under the user's own IAM permissions.
Every step runs with your own IAM permissions.
- Admins can screen MCP traffic with Model Armor and scope agents with VPC Service Controls.
Admins can also govern MCP access with Identity and Access Management (IAM), screen MCP traffic with Model Armor, and scope agents with VPC Service Controls and Principal Access Boundary policies.