preloader

· devops aws ai development

AWS Kiro Web Enters Public Preview: Spec-Driven AI Development Now Runs in the Browser

Source: AWS Blog

AWS announced this week that Kiro Web, the browser-based interface for its spec-driven AI development environment, has entered public preview for all Kiro subscribers with no waitlist required. The addition of a browser interface expands how teams can access Kiro’s capabilities beyond the desktop IDE, and introduces the only current surface where Kiro’s fully autonomous development mode is available.

What Kiro does and why the browser interface is notable

Kiro is AWS’s agentic IDE, built on a VS Code fork with a deep agent layer integrated throughout the development workflow. Its distinguishing characteristic compared to other AI coding tools is spec-driven development: rather than jumping directly to code generation, Kiro first generates a requirements document and acceptance criteria based on a description of what you want to build. Code is written against that spec, which gives the agent a structured target and reduces the tendency for AI-generated code to drift from intent as a task grows in complexity.

The browser interface, Kiro Web, makes all of this available without a local install. It runs in cloud sandboxes and is the current home for autonomous mode, where Kiro can work through tasks independently rather than requiring a developer to stay present and approve each step.

This is a meaningful capability gap to fill. Spec-driven development with an autonomous agent that can work while you are not at your desk has been the most-requested Kiro capability since the tool launched, and its initial availability in the browser interface rather than the desktop IDE suggests that cloud sandboxing is the architecture AWS chose to support that mode safely.

Replacing Amazon Q Developer

Kiro replaces Amazon Q Developer, which AWS announced will reach end of support on April 30, 2027. New Amazon Q Developer sign-ups were blocked starting May 15, 2026. Teams currently using Q Developer have roughly twelve months to migrate workflows to Kiro, but given how quickly Kiro’s feature surface is expanding, moving sooner allows teams to take advantage of the more capable spec-driven approach rather than just replicating what Q Developer offered.

Kiro is available at a free tier of 50 interactions per month, with the Pro plan at $19 per month for higher usage. During the current preview period, Kiro Web consumes credits at half the standard rate.

Kiro in the context of AWS infrastructure work

For teams running workloads on AWS, the integration angle is worth paying attention to. Kiro can use the AWS MCP Server, which reached general availability earlier in May, to access every AWS service with proper authentication from within the IDE. That means infrastructure-related development tasks – writing CDK stacks, modifying Lambda functions, adjusting ECS task definitions – can be handled by an agent with live access to service APIs rather than working from static documentation.

For engineering teams managing significant AWS footprints, this combination of spec-driven development and live AWS service integration is a material change in how infrastructure code can be developed and reviewed.

What to consider before adopting

Kiro Web is in preview and currently limited to the US East (N. Virginia) region. Teams with data residency requirements outside the US should wait for additional regional availability before using it with sensitive codebases. The autonomous mode in particular requires careful thought about what permissions the agent is given in cloud sandbox environments before it runs without supervision.

The spec-generation step adds a stage to the development workflow that teams used to prompt-and-accept AI coding will need to adjust to. The investment is worthwhile for complex features where requirements clarity reduces rework, but may feel like overhead for simple, well-understood tasks.

If you want help evaluating whether Kiro fits your team’s AWS development workflow, migrating from Amazon Q Developer, or setting up a governed AI development environment on AWS, contact Excello Digital. We help teams adopt new AWS tooling in ways that fit their existing processes and security requirements.

We’ll help you resolve your infrastructure challenges

Our team of experts is ready to help you with your infrastructure challenges. We’ll give you honest and personal treatment. Get in touch to learn more.

Get in touch!