Open sourceTypeScript · ≈31k stars · Apache-2.0
Qwen Code
Alibaba's official terminal coding agent.
Qwen Code is Alibaba's official terminal coding agent — forked from Gemini CLI's agent loop and retuned for Qwen models, with the same sandboxed tool-call harness as its upstream. The default harness entry point for the Qwen ecosystem.
Activity
- Last push
- · today
- Latest release
- desktop-v0.28.0
· 4d ago - Last 90 days
- 20+ releases
v0.28.0-nightly.20261120.19b31401f5Release v0.28.0-nightly.20261120.19b31401f5Pre-release
· yesterdayWhat's Changed test(managed-agent): observe a real scheduled /loop fire in the full-stack E2E by @wenshao in https://github.com/QwenLM/qwen-code/pull/17506 fix(verify-pr): harden the watch-state shape guard, container nonce, and VM prose by @wenshao in…
v0.28.1-preview.1Release v0.28.1-preview.1Pre-release
· yesterdayWhat's Changed test(managed-agent): observe a real scheduled /loop fire in the full-stack E2E by @wenshao in https://github.com/QwenLM/qwen-code/pull/17506 fix(verify-pr): harden the watch-state shape guard, container nonce, and VM prose by @wenshao in…
v0.28.0-nightly.20261118.d9ac287c9fRelease v0.28.0-nightly.20261118.d9ac287c9fPre-release
· 3d agoWhat's Changed test(managed-agent): observe a real scheduled /loop fire in the full-stack E2E by @wenshao in https://github.com/QwenLM/qwen-code/pull/17506 fix(verify-pr): harden the watch-state shape guard, container nonce, and VM prose by @wenshao in…
v0.28.1-preview.0Release v0.28.1-preview.0Pre-release
· 4d agoWhat's Changed test(managed-agent): observe a real scheduled /loop fire in the full-stack E2E by @wenshao in https://github.com/QwenLM/qwen-code/pull/17506 fix(verify-pr): harden the watch-state shape guard, container nonce, and VM prose by @wenshao in…
desktop-v0.28.0Qwen Code Desktop v0.28.0
· 4d agoWhat's Changed feat(web-shell): add sidebar workspace filter by @callmeYe in https://github.com/QwenLM/qwen-code/pull/17244 feat(managed-agent): add the DingTalk adapter on the managed channel path by @wenshao in https://github.com/QwenLM/qwen-code/pull/17136…
Checked against GitHub on Nov 21, 2026. Release notes are shortened to plain text.
In research (6)
- Hidden in Plain Sight: an Empirical Study about The Instruction Stacks of Coding AgentsVerdict: Important
Oct 27, 2026 · Captured agent; tool definitions take 72.3% of its latest request.
- A Unified Arena for LLM Agents across Diverse EurogameVerdict: Useful
Oct 26, 2026 · Native harness that gives Qwen3.8-Max a +29.7-point edge over OpenCode on public-spatial games.
Oct 26, 2026 · Harness in Table 12; gives Qwen3.8-27B its worst Repository Editing score.
- ProductWebBench: Evaluating Website Continuity Beyond Task CompletionVerdict: Important
Oct 22, 2026 · Evaluated scaffold on Claude Sonnet 4.6.
- FrontierCS 2: A Benchmark for Computer Science R&DVerdict: Important
Oct 21, 2026 · Harness for Qwen 3.8 Max.
Oct 21, 2026 · Role-typed harness, 0% laundering.
Neighbours on the workbench
Shell · ≈329k stars · MIT
mattpocock/skills
Small, composable agent skills for real engineering work.
Maintained by Matt Pocock (AI Hero)
MDX · ≈323k stars · MIT
obra/superpowers
An agentic skills framework and software methodology that works.
Maintained by Jesse Vincent (obra)
JavaScript · ≈308k stars · MIT
affaan-m/ECC
Agent harness performance optimization system: skills, instincts, memory and security for coding agents.
Maintained by Affaan Mustafa
Connections
Relations from the Harness Hub knowledge graph, as charted in the State of Harness reports.
