AWS releases a free local emulator written entirely by AI agents

Amazon Web Services (AWS) on November 11 released AWS Emulator, a tool that simulates AWS services on a developer's own machine. Its headline feature is how it was made: every line of code was written by autonomous AI agents.

"The emulator was written entirely by autonomous agents. Our team hasn't written or reviewed any of the code," wrote Matt Goldberg, the AWS senior principal engineer behind the project.

What it does

According to the AWS News Blog, the emulator is built from more than 400 API models covering more than 200 services. The AWS CLI, SDKs and coding agents talk to it over a local endpoint using the same APIs they use in the cloud, so developers can test code offline and without an AWS account. It carries no service fee, and its source is on GitHub.

GeekWire noted that the tool targets a problem coding agents such as Kiro and Claude Code have made worse: agents write code faster than teams can test it against the real cloud, where repeated test runs add delay and cost.

Three teams of agents

Amazon says the emulator is built by three teams of agents. One writes "fidelity tests" checked against real AWS, a second runs those tests against the emulator and files issues for failures, and a third fixes each issue. Every change must pass a merge gate in which other agents review coding standards, security and regressions. Goldberg said hundreds of agents now fix thousands of issues a day.

Limits

IDC analyst Devin Pratt told InfoWorld that the emulator does not yet simulate throttling, outages or regional differences, and does not sandbox the code it runs. The agents keep a public list of known gaps in a GitHub folder called "things-that-do-not-work." The New Stack reported that AWS will not accept outside pull requests.

Sources: AWS News Blog, About Amazon, GeekWire, The New Stack