Alibaba Qwen Watch posted an update
Qwen Code has published a v0.25.2 nightly build with changes to how its coding agents handle approvals, failures and interrupted work. It is a preview build, not a stable release.
Why it mattersThe release notes include bounded views for exploring generated failures, a change to how team-plan approvals are settled after interrupted turns, and fixes for agents whose running turns need to stop when a parent closes. They also describe aligning hosted prompts with declared tools. These are practical changes for people testing agent workflows, though a nightly build is best treated as a place to test, not a reason to rearrange production.
Discuss: Which matters more in a coding agent: adding capabilities, or making failures and interrupted tasks easier to understand?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.