Building AI agents has become dramatically easier over the past year, but many development teams still spend too much time on project setup instead of solving business problems.
Traditionally, creating an agent with Google’s Agent Development Kit (ADK) meant building the agent itself, then separately configuring project scaffolding, evaluation workflows, deployment pipelines, CI/CD, observability, and supporting infrastructure. While these steps are important, they often create unnecessary overhead during the early stages of a project.
A newer workflow is changing that.
By combining Google’s Antigravity CLI with Agents CLI, developers can generate a production ready ADK project in a single conversational prompt, a workflow often referred to as one shotting an agent. Instead of manually assembling every component, AI generates the complete project structure, allowing teams to move from idea to working prototype significantly faster.
What Does “One Shotting” Mean?
One-shotting is the process of describing an entire application in natural language and allowing an AI coding assistant to generate nearly everything required in a single pass.
Rather than writing dozens of setup commands, developers can simply describe:
- The business problem
- The agent’s responsibilities
- Required tools and APIs
- Desired workflows
- Deployment requirements
The AI then builds the project structure, code, configuration files, and supporting assets with minimal manual intervention.
The result isn’t necessarily a finished production system but it’s often an excellent starting point that would normally take hours or days to assemble.
The Role of Antigravity
Antigravity serves as Google’s agentic coding environment, allowing developers to build software using natural language instead of manually creating every file. It supports parallel subagents that can research, generate code, validate outputs, and perform multiple development tasks simultaneously.
Instead of telling an assistant what file to create next, developers describe the desired outcome.
For example:
“Build an ADK customer support agent that searches internal documentation, summarizes responses, logs conversations, includes deployment configuration, and prepares evaluation tests.”
Rather than generating only Python code, the Google Antigravity environment can coordinate the broader implementation workflow.
Where Agents CLI Fits In
While Antigravity focuses on intelligent code generation, Agents CLI provides the operational tooling around Google’s Agent Development Kit.
It understands the complete ADK lifecycle, including:
- Project scaffolding
- Local development
- Evaluation workflows
- Deployment
- CI/CD configuration
- Observability
- Agent publishing
Instead of learning numerous commands and best practices manually, developers can instruct their coding assistant to use Agents CLI skills throughout the project. Detailed documentation on how to initialize this workflow can be found in the Google Cloud Agent Platform Guide. The tooling guides the agent toward recommended project structures and deployment patterns.
Eliminating Boilerplate
One of the biggest productivity gains comes from removing repetitive setup work.
Many AI projects begin with the same tasks:
- Creating directory structures
- Configuring environments
- Writing deployment files
- Establishing evaluation frameworks
- Connecting authentication
- Preparing CI/CD
- Organizing reusable components
These tasks are necessary but rarely differentiate one project from another.
By allowing Antigravity and Agents CLI to generate this foundation automatically, developers can spend more time refining business logic, improving prompts, integrating enterprise data, and testing real world workflows.
A Better Starting Point for ADK Projects
The workflow demonstrated in Google’s recent AI agent development examples highlights an important shift in how teams can approach ADK projects.
Rather than treating infrastructure as a prerequisite, infrastructure becomes part of the generated output.
A typical workflow now looks something like this:
- Describe the desired agent in natural language.
- Antigravity plans and generates the project.
- Agents CLI applies ADK best practices for scaffolding, evaluation, and deployment.
- Developers review, refine, and extend the generated solution.
- Iterate on business capabilities instead of rebuilding project foundations.
The emphasis shifts from writing setup code to improving agent behavior.
Why This Matters for Enterprise Teams
Organizations experimenting with AI agents often need to build many prototypes before identifying production ready solutions.
Reducing setup time means teams can:
- Validate ideas more quickly
- Build proof of concepts faster
- Standardize project structures
- Reduce onboarding time for new developers
- Encourage experimentation without large engineering investments
For consulting organizations like HiView Solutions, this translates directly into faster client engagements and shorter implementation timelines.
Instead of spending valuable project hours creating the same infrastructure repeatedly, consultants can focus on understanding business processes, integrating enterprise data, and delivering measurable outcomes.
The Future of Agent Development
Agent development is evolving beyond simply writing Python code.
Modern AI tooling increasingly handles:
- Project generation
- Architecture recommendations
- Deployment preparation
- Evaluation pipelines
- Operational best practices
Developers remain responsible for validating outputs, refining workflows, ensuring security, and designing exceptional user experiences but much of the repetitive engineering work can now be automated.
The combination of Antigravity and Agents CLI represents an important step toward that future.
Final Thoughts
As Google’s AI development ecosystem continues to mature, workflows that combine Antigravity with Agents CLI are likely to become the default starting point for many ADK projects.
By collapsing scaffolding, deployment preparation, evaluation tooling, and operational setup into a single conversational workflow, development teams can focus on building intelligent solutions instead of repetitive project infrastructure.
For organizations building enterprise AI agents, this means faster prototypes, more consistent delivery, and a shorter path from concept to production.
At HiView Solutions, we’re closely following these advancements as we continue helping organizations adopt Google AI technologies and build practical, scalable agent solutions.