Qwen Code v0.26.3 ships workflow mode, memory cadence and compaction fixes
qwen-code-review-bot · ghdev · 2026-10-28
QwenLM/qwen-code released v0.26.3. Highlights: opt-in memory extraction cadence after a no-op run; core fixes for recovering outer XML calls with quoted content and shrinking tool results to the headroom left before auto-compaction; a new persistent session Workflow mode (DW-04A); web-shell now retains Generated UI instances across same-page navigation and adds a configurable simple composer; CLI keeps autofix-pr off the user's own branch work and honors channel approval mode without webhooks; managed-agent queue fixes, ACP rewind identity passthrough, plus CI and release-process fixes.
More from coding & agent
- Anthropic models burn 2-3x tokens and cost 10-50x more on FrontierCode 2.0 — teortaxesTex · 2026-10-28
- What to learn in the AI coding era: the bottleneck moved from writing to verifying code — irombie · 2026-10-28
- Devs praise OpenAI's Decisions API: Codex integration drives full-time switch — craigsdennis · 2026-10-28
- Personal AI agent trains a chess transformer on dual RTX 4090 and battles Sunfish — andreisavu · 2026-10-28
- Opus 5.5 ultracode workflow with 9 subagents builds a Three.js scene — nptacek · 2026-10-28
- Agent sandboxes talk at PLATFORMday SF: the missing runtime for AI — gethackteam · 2026-10-28