AI BriefingGoogle unveils agents-cli
Key point
Google has released agents-cli, which injects ADK development, evaluation, and deployment knowledge into coding agents.
Details
Google's agents-cli is a CLI and skill collection that provides Google Cloud agent development knowledge and task tools to coding assistants such as Claude Code, Codex, and Antigravity CLI.
It supports the development lifecycle of ADK-based agents as follows.
- scaffold: create, extend, and upgrade agent projects
- run: run agents locally
- eval: generate reasoning traces, scoring, dataset synthesis, failure analysis, prompt optimization
- deploy: deployment to Agent Runtime, Cloud Run, GKE, and CI/CD
- publish: registration with Gemini Enterprise
- observability: integration with Cloud Trace, logging, and third-party observability
Installation is done via uvx google-agents-cli setup, and for local development and evaluation, an AI Studio API key can be used without a Google Cloud account. Cloud deployment and operations require the user's own Google Cloud project and associated costs.
A total of 7 skills inject domain knowledge into coding agents, covering the ADK Python API, evaluation methodology, deployment and infrastructure, and observability. It primarily targets teams looking to develop, evaluate, and deploy ADK agents on Google Cloud.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Foremerge v0.5.0 Released: Rust-Based Open-Source Tool Detects Intent Conflicts Between Parallel AI Coding Agents