-
AWS Open Sources Kiro Crew for Asynchronous Coding Agents
Amazon open-sourced Kiro Crew, a system for running multiple Kiro coding agents asynchronously across sessions, handling tasks like incident investigation, ticket triage, and PR monitoring without supervision. Developed internally as MeshClaw and reportedly used by more than 39,000 developers, it ships with an OS-level sandbox, credential redaction, and a signed audit log, and is released under Apache 2.0 for macOS, Linux, and Windows.
# -
Z.ai’s GLM-5.3 goes open weight, but its new license aims at hyperscalers
Z.ai released the weights for GLM-5.3 on Hugging Face, two weeks after launch, following additional safety evaluation. The new license, unlike GLM-5.2's MIT license, requires companies with over $10 billion in revenue across any 12 months to pass a Z.ai security review before commercial hosting. Z.ai reports the model scored 84.5 percent on the CyberGym vulnerability benchmark and found 2,436 vulnerabilities across 269 open-source projects, including the Linux kernel.
# -
When you keep AI Lean, you keep AI correct
Ryan chats with Leo de Moura, Senior Principal Applied Scientist at AWS and the creator of the Lean language, about proving correctness in AI agents with the Lean language, how automated reasoning complements probabilistic AI models, and the use of AI for continuous code optimization.
# -
How to Make Testing More Sustainable
At OOP Conference, Jochen Joswig outlined a sustainable testing strategy: skip unnecessary tests, fail fast so later tests are skipped after an early failure, and run only the tests affected by a given code change. He also recommended tracking energy consumption per test and using static analysis tools like Creedengo, a SonarQube plugin supporting Java, Python, JavaScript, and C#, to catch inefficiencies such as database queries run inside loops.
#