Google Cloud Data Agent Kit is now Generally Available, adding BigQuery Graph, Bigtable, and Apache Spark support

General Availability of Data Agent Kit

Google Cloud has announced the general availability of its Data Agent Kit. This kit is a free collection of Model Context Protocol (MCP) tools and agent skills designed to enable coding agents to work directly with Google Cloud data products. It supports various popular coding agents, including Antigravity, Claude Code, and Codex.

Expanded Data Product Support

With this general availability release, Data Agent Kit now includes support for BigQuery Graph, Bigtable, and Managed Service for Apache Spark. These additions allow agents to access and manage data within open Lakehouse environments. The update also incorporates numerous quality-of-life improvements to enhance daily workflows.

Bridging the Context Gap for Coding Agents

Coding agents are proficient in generating SQL, PySpark, and pipeline code but often lack contextual information about the user's data environment, such as table structures or job statuses. Data Agent Kit addresses this by providing MCP tools that connect to over 15 Google Data Cloud services, enabling agents to inspect schemas, execute queries, read job logs, and manage live resources. Additionally, it offers Google-authored skills, which are open-source instructions from Google Cloud engineers that guide agents in applying data best practices, such as optimizing BigQuery SQL or designing Bigtable row keys.

Integration and Workflow

Data Agent Kit can be used across various development environments. It is available as an IDE extension for VS Code, Antigravity IDE, Cursor, and other VS Code-compatible editors. It also functions as a plugin for Antigravity 2.0, Antigravity CLI, Claude Code, and Codex. The kit comes pre-installed in Cloud Shell and Cloud Workstations. The IDE extension integrates a lightweight version of the Google Cloud console, allowing users to browse data, run queries, and review agent work without switching applications.

Example Use Case

For instance, if a user asks an agent to forecast demand and check inventory, Data Agent Kit loads relevant skills to follow Google's best practices. The agent can then use Knowledge Catalog to identify trusted sales and inventory tables, employ MCP tools to run a forecast in BigQuery, check stock levels in AlloyDB for PostgreSQL, and present the combined results within the user's editor or terminal.