Language: 日本語English

spanner Plugin

Author
Google LLC
Category
Database
Topics
Databases & Storage · AI Agents & AI App Development
First seen
2026-07-09
Last confirmed
2026-07-13
Explanation last updated
2026-07-12

The explanation below is AI-generated. Please verify it against the sources.

Spanner is a set of agent skills (currently in beta, pre-v1.0) that lets AI agents such as Antigravity, Claude Code, and Codex connect to Google Cloud Spanner and let users explore schemas, run queries, and manage databases using natural language. According to the README, it requires a Google Cloud project with the Spanner API enabled, Application Default Credentials, and the IAM roles Cloud Spanner Database Reader and Cloud Spanner Database User. Configuration is done via environment variables (SPANNER_PROJECT, SPANNER_INSTANCE, SPANNER_DATABASE, and the optional SPANNER_DIALECT). Installation steps differ by agent: Antigravity (CLI or 2.0/IDE), Claude Code, and Codex each have their own install commands, and skills can also be installed via the npx skills tool. The repository is licensed under Apache-2.0 and currently provides one documented skill, 'Spanner Data', for exploring schema objects (tables, graphs) and executing custom SQL queries.

Overview

According to the source, this is not a SaaS product but a GitHub repository providing 'agent skills' — natural-language interaction modules that plug into AI agents (Antigravity, Claude Code, Codex) so they can connect to and operate on Google Cloud Spanner databases.

What you can do with spanner

  • Explore database structure: list tables and discover schema objects such as tables and graphs
  • Query data using natural language prompts (e.g., "Show me all tables in the 'orders' database")
  • Run custom SQL queries against a Spanner database via the agent
  • Generate code, such as a Python dataclass, based on a table's schema
  • Install and configure the skill for Antigravity, Claude Code, or Codex using agent-specific commands
  • Set connection details via environment variables (SPANNER_PROJECT, SPANNER_INSTANCE, SPANNER_DATABASE, SPANNER_DIALECT)
  • Use the agent's debug mode (e.g., gemini --debug) to troubleshoot connection or credential issues

Sources

Original description (English)

Connect and interact with Spanner data using natural language.

History of spanner

Back to list