Google Cloud Releases Agents CLI to Speed AI Agent Development
Google Cloud has launched Agents CLI, a command-line tool designed to reduce friction in building AI agents from prototyping through production deployment. The tool addresses a persistent problem: developers juggle disconnected services, documentation, and environments when moving agent projects from experimental to live systems.
Agents CLI integrates with coding assistants like Gemini CLI, Claude Code, and Cursor, creating a single interface to Google Cloud services including Agent Platform, Cloud Run, and infrastructure automation. The goal is straightforward-eliminate the context switching that slows development cycles.
How it works
The CLI embeds structured knowledge about Google Cloud services directly into the tool. Instead of coding agents having to infer how services connect, they access predefined "skills" and API references. This reduces token usage and iteration cycles by making interactions more deterministic.
Developers can initialize projects, define workflows, and configure deployments through a series of commands. Manual setup and extensive documentation reading become optional.
Testing before deployment
Agents CLI includes local simulation and evaluation features. Developers run evaluation pipelines, compare outputs across runs, and validate agent behavior against test datasets before pushing to production.
This reflects a broader shift in agent development toward testing rigor. As agents handle more consequential tasks, accuracy and consistency requirements tighten.
Deployment and control
The tool automates infrastructure provisioning and CI/CD pipeline setup. It generates Infrastructure as Code, deploys agents to Cloud Run or Kubernetes, and publishes to enterprise environments including Gemini Enterprise.
A "Human Mode" feature lets developers execute commands directly instead of relying entirely on agent-driven automation. This addresses concerns about transparency in fully autonomous systems by providing inspection and control points.
Getting started
Agents CLI is available through a standard installation command and includes built-in workflows for project creation, evaluation, and deployment. Google Cloud has published documentation and a GitHub repository to support adoption.
For developers working with AI coding tools or building systems with large language models, the release signals where cloud vendors are investing-reducing the gap between local experimentation and production systems.
Your membership also unlocks:
