Google Releases 'gws' CLI for Workspace Agents
Google has released 'gws,' a new command-line tool to unify Gmail, Drive, Calendar, and Docs APIs for building AI agents. The tool allows developers to more easily orchestrate complex, automated workflows and agentic pipelines across the entire Google Workspace suite from a single interface.
The new 'gws' command-line tool is built in Rust and distributed via npm, offering a single interface for nearly all Google Workspace APIs, including Drive, Gmail, Calendar, Sheets, Docs, and Chat. It gained immediate traction with developers, accumulating 4,900 GitHub stars within three days of its March 2, 2026, release. Unlike traditional CLIs with fixed commands, 'gws' dynamically generates its commands at runtime by reading Google's Discovery Service. This means that when Google adds new API endpoints, the CLI supports them automatically without requiring a software update. All output is in structured JSON, making it easily parsable for automated workflows. The tool is explicitly designed for AI agents, shipping with over 100 pre-built "agent skills" and more than 50 curated recipes for automating common tasks. These skills allow agents to perform actions like uploading files to Drive, sending emails, or creating calendar events. The documentation specifically highlights integration with OpenClaw, a popular open-source platform for AI agents. 'gws' includes native support for the Model Context Protocol (MCP), an open standard for AI agents to interact with external tools. By running as an MCP server, it allows any compatible AI client, such as Anthropic's Claude, OpenAI's ChatGPT, or Google's own Gemini, to use Workspace APIs as structured tools. This aligns with a broader industry trend of enabling AI agents to access and control business software. While presented as a developer example and not an officially supported product, the tool's release signals a larger strategic move by Google. It positions Google Workspace as an accessible platform for a growing ecosystem of autonomous AI systems, competing with similar initiatives like Microsoft's Copilot. For developers, 'gws' can be installed via npm. Authentication requires setting up a Google Cloud project and uses OAuth, with credentials stored securely in the operating system's keyring. Although the initial setup can be complex, requiring the gcloud CLI, it offers various authentication flows for different environments, including headless machines and server-to-server interactions.
