Back to Skills

google-agents-cli-workflow

This skill should be used when the user wants to "develop an agent", "build an agent using ADK", "run the agent locally", "debug agent code", "test an agent", "deploy an agent", "publish an agent", "monitor an agent", or needs the ADK (Agent Development Kit) development lifecycle and coding guidelines. Entrypoint for building ADK agents. Always active — provides the full workflow (scaffold, build, evaluate, deploy, publish, observe), code preservation rules, model selection guidance, and trouble

2,079stars246forksUpdated 5/6/2026

Security Assessment

Low Risk(80/100)

Detected risks:

Sensitive File Access([SKILL.md] .env)
Security Score80/100

About google-agents-cli-workflow

The `google-agents-cli-workflow` skill is designed to guide users through the complete lifecycle of agent development using the Agent Development Kit (ADK). It provides a structured workflow for creating, building, testing, deploying, publishing, and monitoring agents, ensuring best practices and coding standards are followed throughout the process. By consolidating the essential steps of agent development into a single, coherent skill, it helps prevent common mistakes such as skipping scaffolding or misconfiguring evaluation and deployment settings, which could lead to errors or incomplete projects.

This skill offers a range of capabilities including scaffolding new projects or enhancing existing codebases, guidance on code evaluation and testing, deployment configuration, and publishing options for Gemini Enterprise. It provides detailed instructions for version management, project structure, session continuity, and cross-skill references for each phase of development. The skill also maps product names to the correct CLI values, addresses installation and setup requirements, and provides troubleshooting steps for issues like missing tools or version mismatches. Users are guided to follow a step-by-step workflow to ensure code quality, model selection, and observability.

Target users include developers and engineers working with Google Cloud's agent platforms, particularly those using Gemini CLI, Claude Code, Codex, or similar coding agents. It is suited for both newcomers who need scaffolding and guidance, as well as experienced developers who want to maintain workflow consistency and best practices. Use cases include building new AI agents from scratch, testing and debugging agent behavior, deploying agents to production environments, and monitoring live agent performance for optimization and reliability purposes.

FAQ

How do I start a new project using this skill?

Use `agents-cli scaffold create <name>` to scaffold a new project. If you already have code, use `agents-cli scaffold enhance .` to integrate the required CLI structure.

Which versions of `google-agents-cli` are compatible?

This skill requires `google-agents-cli` version 0.1.2 or higher. You can check your version with `agents-cli info` and install or update with `uv tool install 'google-agents-cli~=0.1.2'`.

Do I need to install any dependencies before using this skill?

Yes, you need the `uv` tool installed first. Follow the [official installation guide](https://docs.astral.sh/uv/getting-started/installation/index.md) if it is not already available.

Can this skill be used with different agent coding platforms?

Yes, it works with any coding agent such as Gemini CLI, Claude Code, or Codex, providing a unified workflow for development and deployment.

What should I do if a product name differs from the CLI value?

Refer to the product name mapping table in the skill documentation to map user terminology to the correct CLI parameters for deployment and datastore configuration.

All Files

2 files
SKILL.md23.2 KB
View
references/internals.md2.2 KB
View

Install google-agents-cli-workflow

Download and extract the skill files to your .claude/skills/ directory.

Quick Setup:

  1. Copy the skill folder to .claude/skills/
  2. Claude will automatically detect and use the skill