Qwen Code v0.25.4-preview ships session-targeted Skill refresh and agent hardening
qwen-code-review-bot · ghdev · 2026-10-21
QwenLM/qwen-code released v0.25.4-preview.1 with over a dozen changes:
- Daemon now supports session-targeted Skill and Extension refresh; managed agents gain local and SSH workspace file references.
- New read-only --check preflight mode and a configuration template shipped in the reference distribution.
- Runtime broker warms session creation for workspace-isolated bound sessions; fixes recovery of committed wakes after writer loss.
- Serve-side hardening: Hosted stdio MCP servers contained in a process group, private CSI file turns routed through Hosted Harness, daemon session creation sized from the session budget.
- Numerous test stability fixes: slow-host turn-settling budgets, hosted probe-start scheduling budgets, ENOTEMPTY retries in teardowns.
More from coding & agent
- Index Network open-sources Social Discovery Protocol: agents negotiate fit on your behalf — carsonfarmer · 2026-10-23
- Installed a formal-verification skill: it silently installed Java, MacBook runs 10 degrees hotter — zealcaiden · 2026-10-23
- OpenAI's Decisions API hits GA: near real-time routing up to 10x faster than GPT-6 Luna — OpenAIDevs · 2026-10-23
- Ruff author warns: upgrade Node/Bun now — --use-system-ca still accepts the eMudhra cert — charliermarsh · 2026-10-23
- Proof of Agent: paste one prompt and your AI mines, grades, and self-governs — adamamcbride · 2026-10-23
- DeepLearning.AI weekly: GPT-6.1 hits 97.9% byte parity on COBOL migration — DeepLearningAI · 2026-10-23