Getting Started with Kiro Crew
AI coding assistants are becoming an essential part of modern software development. But most of them still require you to stay in the loop: ask a question, wait for the response, provide another instruction, and repeat.
AI coding assistants are becoming an essential part of modern software development. But most of them still require you to stay in the loop: ask a question, wait for the response, provide another instruction, and repeat.
Kiro Crew takes a different approach.
Crew is a persistent AI development workspace that can maintain context, run long-running tasks, work on multiple tasks in parallel, use project knowledge, and run scheduled jobs.
In this post, we'll get started with Kiro Crew and explore some of its key features.
What is Kiro Crew?
Kiro Crew is built around the idea of having an AI development assistant that can keep working with you over time.
Some of the capabilities include:
- Persistent sessions and memory
- Project knowledge
- Parallel tasks
- Scheduled jobs
- Long-running development tasks
- Reusable skills
- Multiple ways to interact with Crew
This makes Crew useful for more than just generating code. It can become part of your development workflow.
Installing Kiro Crew
Start by installing the Kiro CLI.
1
curl -fsSL https://cli.kiro.dev/install | bash
Once installed, launch Crew:
1
kiro-cli crew
Follow the authentication steps to connect your Kiro account.
After signing in, you can open the Crew interface and start your first session.
Your first task
Let's start with a simple repository analysis.
For example:
1
2
3
4
Analyze this repository and explain its main components,
architecture, and testing strategy.
Don't modify any files.
Crew can inspect the project and provide an overview before you start making changes.
You can then continue with a more specific task:
1
2
Identify the areas of the project with the lowest test coverage
and create an implementation plan to improve them.
Using project knowledge
Large projects contain a lot of documentation and context.
Crew can use project knowledge to make relevant information available during development.
For example, you can provide architecture documentation and then ask:
1
2
Based on the project documentation,
explain how authentication works in this application.
This can be particularly useful for large repositories where understanding the existing architecture is just as important as writing new code.
Enjoyed reading this content? Let the author know!
Your likes, comments, shares, and saves help creators reach more builders.
Loading recommendations
Loading article