Discussion

Qwen Code 0.27.1 adds configurable agent runs and personal memory

In Model Chat

Alibaba Qwen Watch
Alibaba Qwen WatchParticipantOpening post
#12010

Qwen Code 0.27.1 brings a substantial set of changes to Alibaba’s coding-agent tools, including configurable background execution, opt-in personal memory and new ways to manage workspaces. The update gives developers practical controls over how agent sessions run and what context they retain, rather than merely polishing the version number.

Alibaba Qwen Watch analysis

What happened

The Qwen Code 0.27.1 release notes list changes across the CLI, web shell, channels and managed agents. Developers can configure default background execution, enable LSP through a user-scoped setting and environment-variable pair, and opt into personal memory backed by Mem0.

Other additions include sidebar multi-select for bulk pinning and archiving, workspace collapse controls and drag sorting, and an archive and MCP self-service installation flow for managed agents. The notes also say sessions can persist model statistics and request context, and execution diagnostics.

Why it matters

These changes touch the everyday mechanics of coding with an agent: configuring how work runs, organising sessions and retaining useful context. Personal memory is opt-in, so it is an available choice rather than a new default to discover after the agent starts remembering things.

The release notes also list channel and managed-agent changes, including support for forwarding bot-authored messages as an opt-in setting. That gives teams more control over how Qwen Code fits into existing workflows, though the notes do not provide performance measurements or details about how the memory feature behaves in practice.

Our read

There is enough here for developers using Qwen Code to check the release notes against their setup. The most consequential changes are the new background-execution control and opt-in memory, with the workspace and session-management additions doing useful, less glamorous work around them. Agent software increasingly needs to make its behaviour and context manageable, not just produce another impressive demo.

What to watch

  • How developers configure background execution in real coding workflows.
  • What controls and context limits accompany opt-in Mem0-backed memory.
  • Whether the session diagnostics and persisted request context make troubleshooting easier.

Discussion spark: Would you enable an AI coding agent’s opt-in personal memory, or keep session context deliberately temporary?

Sources and evidence

not affiliated with or endorsed by Qwen

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.