AWS launches Kiro IDE for real agentic development at scale

Specs & hooks

Kiro specs are artifacts that prove useful when developers need to think through a feature in-depth, refactor work that needs upfront planning, or when they want to understand the behaviour of systems.

Because requirements are usually uncertain when developers start building, developers use specs for planning and clarity. Specs can guide AI agents to a better implementation in the same way.

Kiro hooks act like an experienced developer, catching things that engineers miss or completing boilerplate tasks in the background as a developer works. 

These event-driven automations trigger an agent to execute a task in the background when developers save, create, delete files, or on a manual trigger.

“Kiro accelerates the spec workflow by making it more integrated with development. It unpacks requirements from a single prompt — type “Add a review system for products” and it generates user stories for viewing, creating, filtering, and rating reviews. Each user story includes EARS (Easy Approach to Requirements Syntax) notation acceptance criteria covering edge cases that developers typically handle when building from basic user stories. This makes your prompt assumptions explicit, so you know Kiro is building what you want,” says AWS.

Kiro generates a design document by analysing a codebase and approved spec requirements. It creates data flow diagrams, TypeScript interfaces, database schemas, and API endpoints. 

This eliminates the lengthy back-and-forth on requirements clarity that typically slows development.

“Kiro generates tasks and sub-tasks, sequences them correctly based on dependencies, and links each to requirements. Each task includes details such as unit tests, integration tests, loading states, mobile responsiveness, and accessibility requirements for implementation. This lets developers check work in steps rather than discovering missing pieces after they think they’re done,” writes AWS, in a technical blog.

Kiro simplifies this entire process by autogenerating the tasks and sub-tasks, sequencing them in the right order, and linking each task back to requirements so nothing falls through the cracks. Kiro has thought of writing unit tests for each task, added loading states, integration tests for the interaction between products and reviews, and responsive design and accessibility.

Execution status

The task interface lets teams trigger tasks one-by-one with a progress indicator showing execution status. Once complete, users can see the completion status inline and audit the work by viewing code diffs and agent execution history.

“Kiro AI IDE and Kiro Powers are two powerful indicators that AWS is dead serious about building up its chops with the development community. Powers are more than automations performed by agents. A power is a transportable unit of capability that carries with it patterns, guardrails and tool access. As more developers and partners build powers, agents will gain expanded roles in the development process. This is how the AWS developer community begins to normalise agent behaviour across tools and creates a path for real agentic development at scale,” said Mitch Ashley, VP and practice lead, Futurum Research

Kiro’s specs stay synced with an evolving codebase. Developers can author code and ask Kiro to update specs or manually update specs to refresh tasks. This solves the common problem where developers stop updating original artifacts during implementation, causing documentation mismatches that complicate future maintenance.

AWS says its vision in this space is to solve the fundamental challenges that make building software products difficult. It wants to do this all the way from ensuring design alignment across teams and resolving conflicting requirements, to eliminating tech debt, bringing rigour to code reviews and preserving institutional knowledge when senior engineers leave.