Build Production Agents Without Leaving Your Code Editor

Most AI projects get stuck in prototype mode. Moving from a local script to a secure production agent usually requires jumping between half a dozen tools, consoles, IAM dashboards, and deployment platforms. Every context switch adds friction, and momentum fades away.

It doesn’t have to be that way. With the Agents CLI and the Gemini Enterprise Agent Platform, you can go through the entire agent lifecycle without ever leaving your coding agent. Think of it as a calm swamp channel where you can paddle straight from idea to deployment, no portages needed.

What We’re Building: An Industry Watch Agent

This tutorial walks you through building a real Industry Watch agent—a sector-intelligence analyst for semiconductor stocks that reconciles what companies say in the press against what they file with the SEC. We’ll cover six stages: Setup, Build, Deploy, Govern, Evaluate, and Publish.

Stage 1: Teach Your Agent Platform Skills

A general-purpose coding agent writes fine Python, but it doesn’t know ADK’s agent classes or how to deploy to a managed runtime. The Agents CLI closes that gap. Install it and run setup:

uvx google-agents-cli setup

That installs lifecycle skills into your coding agent: scaffolding, deployment, evaluation, and publishing. Add the Developer Knowledge MCP so the agent can look up current platform docs instead of relying on stale training data.

Stage 2: Build the Agent from a Prompt

You won’t hand-write any of this. Describe the agent, and the coding agent scaffolds it. For example, prompt:

"Scaffold a new ADK agent called industry-watch in prototype mode: a sector-intelligence analyst for NVDA, AMD, INTC, MU, and AVGO. Project structure only, no tools yet."

Then add three deterministic FunctionTools: fetch_company_disclosures (SEC EDGAR 8-K filings), fetch_public_claims (GDELT news plus IR feeds), and reconcile_claims_vs_disclosures (join on CIK/ticker and date window, bucket into matched, filing-only, and claim-only, score materiality). The coding agent writes tools.py with typed Python functions that ADK turns into schemas.

Stage 3: Deploy to a Managed Runtime

A local prototype isn’t a service. Deploy to Agent Runtime for a managed, autoscaling home with fast cold starts. Add Sessions for multi-turn state and Memory Bank so the agent remembers your watch-list across weeks. Move deterministic work into the code-execution sandbox to keep it isolated from the model.

Stage 4: Govern and Secure the Agent

Governance is where prompt-driven work usually breaks down. Start with Agent Identity for least-privilege roles. Register the agent in Agent Registry and route traffic through Agent Gateway with an egress allow-list. Then add Model Armor to screen prompts, model responses, and untrusted tool output for prompt injection and jailbreak attempts.

Stage 5: Evaluate Quality with Grounded Evaluations

You can’t ship on vibes. Synthesize a multi-turn eval set and grade with task success, tool-use quality, and hallucination. Add a deterministic metric: every accession number and 8-K item code the agent cites must appear verbatim in tool output. Cluster failures into modes, optimize the prompt against prompt-driven failures only, and prove no regression before keeping the change.

Stage 6: Publish to Gemini Enterprise

An agent someone has to SSH into is an agent nobody uses. Publish the deployed agent to your Gemini Enterprise app using ADK registration. Now the analyst asks, in the same app they use for everything else: “What materially changed for my semiconductor watch-list this week, and which company announcements aren’t backed by an SEC filing?” The answer comes back grounded and cited.

What Comes Next

None of this required a new UI, a second mental model, or a handoff between tools. ADK is open source, the platform services are managed, and the Agents CLI is the connective tissue that lets one assistant drive both. You moved through build, deploy, govern, optimize, and publish in plain English, and stayed in your coding agent the whole time.

Get started with the Agents CLI and build your first agent from a single prompt. The ADK docs cover tools, sessions, and evaluation when you want to go deeper. Your coding agent isn’t just where you write agent code—it’s the control plane for the whole lifecycle.

Original announcement published on Google Cloud.