Kiro Crew

Kiro Crew is a self-hosted AI development workspace for developers who want agents to keep working between conversations. It runs on your computer or a server you control, but its default kiro-cli backend requires separate Kiro sign-in for model access. The default Kiro connection sends model requests to Kiro's cloud service, powered by Amazon Bedrock.

Sessions and task checkpoints survive restarts. Agents can plan multi-step work, check results, retry failures and resume interrupted tasks. Scheduled jobs, monitoring and messaging events can start work while you're away; scheduled scripts and commands can also run without model calls.

Corrections and failures become saved lessons, including lessons limited to a particular repository. Repeated work can become reusable skills. You can inspect and edit both, rather than relying on memory you can't review. Isolated subagents handle parallel assignments, while MCP tools and installable Apps extend the workspace.

The desktop app, browser dashboard and CLI share the same work, with Slack and Discord connections available. The dashboard exposes tool activity, approvals and logs. Execution controls include credential guards, audit records and OS sandboxing where available.

The project is open source under Apache 2.0 and supports macOS, Windows, Linux and Docker. State and memory remain on the Gateway host. Screenshots submitted for diagnosis go to the active Kiro model. Anonymous usage telemetry is enabled by default and can be disabled.