Amazon is adding another tool to the crowded AI coding market with Kiro, a new development environment built to help developers turn software ideas into production-ready work.
The pitch is not just faster code generation. Kiro is designed around a more structured workflow: define the requirement, translate it into technical plans, then break the work into trackable development tasks that stay connected to the original goal.
From prompt to software plan
Kiro starts with what Amazon calls "specs". A developer can enter a prompt such as "Add a review system for products", and Kiro turns that request into user stories.
Those stories cover functions such as displaying reviews, creating reviews and evaluating reviews. Each story also includes acceptance criteria, including edge cases that developers would normally need to identify and manage by hand.
This matters because AI coding tools often move quickly from instruction to generated code. Kiro’s approach puts a planning layer in between. The goal is to make the system capture what the software should do before implementation begins.
That does not remove the developer from the process. Instead, it gives the developer a structured set of requirements to work from, with the AI environment handling much of the early breakdown that usually sits between a product idea and engineering execution.
How Kiro reads the codebase
After the requirement stage, Kiro scans the existing codebase. It then creates technical design documents that may include data flow diagrams, TypeScript interfaces and database schemas.
That step is important because production software is rarely built in isolation. New features have to fit into existing architecture, data models and interface patterns. By reading the current codebase, Kiro can generate plans that are tied to the project already in front of the developer.
The tool then divides the work into development tasks. Those tasks are sorted by dependency, so the order of work reflects what must be built first. Each task is also linked back to the original requirements.
According to the source article, these tasks can include practical implementation details such as:
- unit tests
- integration tests
- loading states
- accessibility
That gives Kiro a broader role than a simple code completion tool. It is trying to organize the whole path from feature request to deliverable code, including the surrounding work that makes software usable and maintainable.
A workspace for tracking implementation
Developers work through Kiro’s tasks in a dedicated interface. That interface tracks progress, displays code diffs and maintains a complete work history.
The source article emphasizes that specs remain synced with the code throughout development. This is a central part of Kiro’s value proposition. Manual documentation often falls behind as code changes. Kiro is meant to reduce that gap by keeping requirements and implementation connected as work moves forward.
For teams, that connection can be useful. If a task changes, the link to the original requirement helps clarify why the code exists and what behavior it is supposed to support. If a code diff appears, the surrounding task history can give reviewers more context than the diff alone.
The result is a workflow where the AI tool is not only producing code. It is also preserving the reasoning trail around the code: what was requested, how it was translated into design, what tasks were created and how implementation progressed.
Hooks add background quality checks
Kiro also includes "hooks", which are designed to help maintain code quality. According to Amazon, these hooks react automatically to events such as saving files and then perform checks or routine work.
AWS describes the hooks as an "experienced developer" working in the background. In practical terms, they can update tests when React components change, keep API documentation current and scan for security vulnerabilities before commits.
Once the rules are configured and pushed to the Git repository, they can apply across the whole team. That makes the workflow less dependent on each individual developer remembering every review step manually.
The source article also notes that Kiro can integrate with external tools through the Model Context Protocol (MCP). That gives the environment a way to connect with other systems rather than remaining limited to its own built-in functions.
Amazon enters a crowded AI coding race
Kiro joins Amazon's Q Developer as part of Amazon’s broader AI coding lineup. Like Cursor, Kiro is built on Microsoft's open source Visual Studio Code platform, which helps it fit into familiar development workflows and support plugins.
The tool is currently free for Mac, Windows and Linux. Access requires joining a waitlist and signing in with Google or Github.
By default, Kiro runs on Claude Sonnet 4 and 3.7. Support for more models is planned. The source article notes that AWS and Claude developer Anthropic have a long-standing partnership that includes billions of dollars in investment from Amazon.
The market around Kiro is already intense. Anysphere, the company behind Cursor, is reportedly valued at $9 billion. Cognition recently acquired rival Windsurf after most of Windsurf's team was "acqui-hired" by Google instead of OpenAI.
Other AI coding tools are also moving quickly, including ByteDance's TRAE, Anthropic's Claude Code, Google's Gemini Code Assist and Microsoft's Github Copilot.
Kiro’s main distinction is its emphasis on structure. Instead of focusing only on producing code from a prompt, it tries to connect requirements, design documents, ordered tasks, tests, documentation and quality checks in one development environment. For Amazon, that is the argument for why another AI coding tool can still stand out in a crowded field.