github QwenLM/qwen-code v0.26.9
Release v0.26.9

one hour ago

What's Changed

  • fix(channels): keep GitHub inbound recovery armed after a failed read by @wenshao in #16402
  • test(core): harden renewal waits and writer seal coverage by @xxwtiancai in #14482
  • feat(verify-pr): give the sandboxed round the PR's review threads by @wenshao in #16300
  • perf(core): read each Managed message body once per process by @wenshao in #16337
  • perf(managed-agent): index the activation lease lookup by @wenshao in #16336
  • perf(managed-agent): narrow SSH settlement and worktree locking by @doudouOUC in #16386
  • fix(core): keep a writer grant an in-flight renewal refreshed by @wenshao in #16287
  • feat(autofix-pr): give helper refusals a code and a next step by @wenshao in #16238
  • test(web-shell): consolidate sidebar suite fixtures on the shared sidebarHarness by @4ekuct25 in #16016
  • docs(design): propose PR state transitions from the daemon PR refresh by @wenshao in #16246
  • perf(core): load the MCP client's Google auth providers on demand by @wenshao in #16242
  • perf(managed-agent): take native journal snapshots off the tenant guards by @doudouOUC in #16410
  • fix(core): keep session list titles when a transcript is rebuilt by @wenshao in #16299
  • feat(qqbot): deliver proactive loop messages as active group messages by @Eric-GoodBoy-Tech in #15635
  • fix(core): accept images for Llama 3.2 Vision, Llama 4, LLaVA, Pixtral and Mistral Small 3.1 and 3.2 by @Shizoqua in #16311
  • feat(cli): have an Ink TUI turn await its shared tool batch by @qqqys in #16385
  • feat(serve): add atomic user Skill activation by @callmeYe in #16186
  • feat(skills): add file-backed archive installation kernel (1/4) by @callmeYe in #16421
  • feat(skills): expose native archive routes and receipts (2/4) by @callmeYe in #16422
  • feat(sdk): support native Skill archives and receipt lookup (3/4) by @callmeYe in #16423
  • perf(managed-agent): take tool execution control off the tenant guards by @doudouOUC in #16415
  • chore(ci): shorten stale PR window to 30 + 14 days by @wenshao in #16435
  • feat(web-shell): polish composer action menu presentation by @ytahdn in #16181
  • fix(ci): apply the apt-mirror probe to the release docker lane too by @yiliang114 in #16278
  • fix(core): make persisted session pagination safe for equal mtimes by @yiliang114 in #16024
  • fix(serve): bound command snapshot replay memory by @doudouOUC in #16413
  • ci: route the no-AK integration gate to ECS for every pull request by @yiliang114 in #16132
  • feat(managed-agent): support durable structured user questions by @doudouOUC in #16433
  • feat(cli): report interactive program status with OSC 7501 by @sorrycc in #16407
  • feat(web-shell): upload Skill archives and reconcile results (4/4) by @callmeYe in #16424
  • feat(channels): journal private bound-session delivery and scan health by @wenshao in #16237
  • fix(web-shell): preserve wrapped drafts during arrow navigation by @jianyx1 in #14909
  • fix(web-shell): batch session deletions at the route limit by @jianyx1 in #14637
  • fix(core): accept images for Qwen VL models under their Ollama names by @Shizoqua in #16309
  • feat(managed-agent): enable automation by default by @doudouOUC in #16448
  • test(managed-agent): cover pending sibling approval holds by @jianyx1 in #15994
  • docs(providers): document OCI Generative AI under OpenAI-compatible providers by @fede-kamel in #16345
  • ci: build the PR tree once and shorten the required path by @sorrycc in #16450
  • fix(subscription): isolate activation and restore sign-in recovery by @liuxuely0626 in #16453

New Contributors

Full Changelog: v0.26.8...v0.26.9

Don't miss a new qwen-code release

NewReleases is sending notifications on new releases.