AIBRIEF
The open source release coordinates multiple AI coding agents across sessions, with persistent memory and audit logs, and runs without an AWS account.
AWS released Kiro Crew on Tuesday as an open-source platform for orchestrating multiple AI coding agents across repositories, sessions, and schedules. The software, which AWS open-sourced to the kirodotdev/KiroCrew repository along with its governance and maintainers files, ships with persistent memory, approval workflows, signed audit logs, and a web/desktop dashboard. Customers can run it on a laptop, container, or VM without an AWS account.
AWS says the project started inside the company as MeshClaw and was adopted by more than 39,000 Amazon builders in less than six months, a figure AWS provided to InfoWorld and that has not been independently audited. AWS ships three reference apps on top of the engine: DevFleets for worktree management, Issue Radar for issue and pull-request triage, and Task Runner for long-running engineering jobs.
Darko Mesaros, a distinguished developer advocate at AWS, called Kiro Crew an application layer for "always-working, self-learning, autonomous teammates," a positioning choice that assumes enterprises are ready to trust multi-agent coding work running unattended. Governance and interoperability remain open questions for enterprise adoption, according to InfoWorld's reporting. Kiro Crew is new enough that no third-party production deployment or independent security review has been published yet.