Amazon Open-Sources Kiro Crew for Asynchronous Coding Agents

System allows developers to assign tasks to AI agents that work across sessions without active supervision

edit
By LineZotpaper
Published
Read Time2 min
Amazon has announced Kiro Crew, an open-source system for running multiple Kiro coding agents across sessions, tools, and tasks. The workspace lets developers assign asynchronous coding tasks to AI agents, enabling work such as incident investigation, ticket triage, migrations, and PR monitoring to continue without active supervision.

Developed internally at Amazon as MeshClaw and reportedly adopted by more than 39,000 developers, Kiro Crew is now available as an open source project. It can run locally or on infrastructure developers control.

Kiro Crew supports persistent multi-session development work through shared memory, reusable skills, scheduled jobs, concurrent agents, and purpose-built Apps. Agents can retain project context, run tasks while developers are away, delegate work to subagents, and integrate with external tools through MCP and webhooks.

According to Amazon engineers Bolin Chen, Zejiang (Joe) Guo, and Zezhen Xu, the project was born from an internal need: "Three of us wanted something simple that wasn't available internally, a way to kick off a task, walk away, come back to something worth reviewing, and to run several tasks at once instead of babysitting one prompt at a time."

The team noted they were inspired by the momentum of OpenClaw and various tools with self-learning agents but needed something that met their security requirements for internal development work.

Kiro Crew orchestrates agents through the Agent Client Protocol (ACP), providing live visibility into their work. An Activity view shows each agent's plan, tool calls, approval gates, and results in real time. The system ships with security features including an OS-level sandbox, denied-by-default commands, suspicious-pattern blocking, input validation, sensitive-path blocking, credential redaction, and a signed audit log of every action.

The system can be customized to fit existing tools and workflows through MCP, Apps, orchestration, and reusable skills. It also supports skills built for other open, standards-based agent platforms without modification.

§

Analysis

Why This Matters

  • Developers gain the ability to offload repetitive coding tasks to AI agents that work asynchronously, potentially increasing productivity
  • Amazon's move to open-source a tool used internally by thousands of developers signals a broader industry trend toward agentic development workflows
  • Security-first design of Kiro Crew may set a precedent for how coding agents handle permissions and auditing in sensitive environments

Background

Kiro Crew is part of a growing ecosystem of AI-powered coding agents that automate software development tasks. Amazon developed it internally as MeshClaw before releasing it as an open-source project. The system uses the Agent Client Protocol, an open standard for agent communication, allowing it to integrate with other tools and platforms. The decision to open-source the tool gives developers control over their infrastructure and data, contrasting with cloud-only agent services.

Key Perspectives

Developers: Gain the ability to run multiple asynchronous coding tasks without active supervision, with features like shared memory, reusable skills, and scheduled jobs. Security teams: The defense-in-depth approach with OS-level sandboxing, command restrictions, and audit logs addresses common concerns about AI agent safety. Critics/Skeptics: The complexity of setting up and managing multiple agents may offset productivity gains for smaller teams or simpler projects. The reliance on agent orchestration could introduce new failure modes.

What to Watch

  • Adoption rate of Kiro Crew in open-source community compared to other agent frameworks
  • Integration with major development platforms like GitHub, GitLab, and CI/CD pipelines
  • Emergence of competitive open-source agent orchestration systems from other cloud providers

Sources

newspaper

Zotpaper

Articles published under the Zotpaper byline are synthesized from multiple source publications by our AI editor and reviewed by our editorial process. Each story combines reporting from credible outlets to give readers a balanced, comprehensive view.