OpenSkill is a framework enabling LLM agents to self-evolve by synthesizing grounded skills and verification signals from open-world resources without external target-task supervision.</p>\n","updatedAt":"2026-06-08T01:55:01.022Z","author":{"_id":"6039478ab3ecf716b1a5fd4d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg","fullname":"taesiri","name":"taesiri","type":"user","isPro":true,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":312,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.9065223336219788},"editors":["taesiri"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.06741","authors":[{"_id":"6a2620e9e4c258a029491fa8","name":"Zhiling Yan","hidden":false},{"_id":"6a2620e9e4c258a029491fa9","name":"Dingjie Song","hidden":false},{"_id":"6a2620e9e4c258a029491faa","name":"Hanrong Zhang","hidden":false},{"_id":"6a2620e9e4c258a029491fab","name":"Wei Liang","hidden":false},{"_id":"6a2620e9e4c258a029491fac","name":"Yuxuan Zhang","hidden":false},{"_id":"6a2620e9e4c258a029491fad","name":"Yutong Dai","hidden":false},{"_id":"6a2620e9e4c258a029491fae","name":"Lifang He","hidden":false},{"_id":"6a2620e9e4c258a029491faf","name":"Philip S. Yu","hidden":false},{"_id":"6a2620e9e4c258a029491fb0","name":"Ran Xu","hidden":false},{"_id":"6a2620e9e4c258a029491fb1","name":"Xiang Li","hidden":false},{"_id":"6a2620e9e4c258a029491fb2","name":"Lichao Sun","hidden":false}],"publishedAt":"2026-06-04T00:00:00.000Z","submittedOnDailyAt":"2026-06-08T00:00:00.000Z","title":"OpenSkill: Open-World Self-Evolution for LLM Agents","submittedOnDailyBy":{"_id":"6039478ab3ecf716b1a5fd4d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg","isPro":true,"fullname":"taesiri","user":"taesiri","type":"user","name":"taesiri"},"summary":"Self-evolving agents requires adaptation after deployment, but existing approaches assume a usable learning loop, such as curated skills, successful trajectories, or verifier signals. Real open-world deployments may provide none of these, offering only a task prompt. In this work, we study open-world self-evolution, where an agent must build both its skills and its own verification signals from scratch, using open-world resources but no target-task supervision. We propose OpenSkill, a framework that bootstraps this loop: it acquires grounded knowledge and verification anchors from documentation, repositories, and the web, synthesizes them into transferable skills, and refines those skills against self-built virtual tasks grounded in the anchors rather than in target answers. The open world thus supplies both the knowledge to be learned and a supervision-independent practice environment, with target-task supervision reserved for final evaluation. Across three benchmarks and two target agents, OpenSkill attains the best automated pass rate while satisfying the no-supervision constraint. Analysis shows its skills transfer across models without model-specific adaptation, and its self-built verifier aligns with ground-truth outcomes despite never accessing them.","upvotes":6,"discussionId":"6a2620e9e4c258a029491fb3","projectPage":"https://github.com/OpenLAIR/OpenSkill","ai_summary":"OpenSkill enables self-evolving agents to develop skills and verification signals from scratch using open-world resources without target-task supervision, achieving high automated performance across benchmarks.","ai_keywords":["self-evolving agents","open-world deployment","skill acquisition","verification signals","bootstrapped learning loop","grounded knowledge","transferable skills","self-built virtual tasks","automated pass rate","skill transfer"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct"},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"619f01b8cc04eadf54fa5d5d","avatarUrl":"/avatars/313e67bf0b13a69a76465dc589f3bbce.svg","isPro":false,"fullname":"Song Dingjie","user":"songdj","type":"user"},{"_id":"65aae63fc3fa44c7109559bb","avatarUrl":"/avatars/b3f3e5d09b410f717c07b6aea997d595.svg","isPro":false,"fullname":"Zhiling Yan","user":"JuelieYann","type":"user"},{"_id":"6675344c20ff491d381a3f1e","avatarUrl":"/avatars/e3ebfce52d4029aa241f40051fcc1e0c.svg","isPro":false,"fullname":"Henry Zhang","user":"Henry0709","type":"user"},{"_id":"6039478ab3ecf716b1a5fd4d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg","isPro":true,"fullname":"taesiri","user":"taesiri","type":"user"},{"_id":"65642d7401de72cb63165d22","avatarUrl":"/avatars/1f4417c4ac5e781ce73eae1060e3f7f2.svg","isPro":false,"fullname":"ytaewon","user":"hamzzi","type":"user"},{"_id":"624ac233c04d55ec0f42b11e","avatarUrl":"/avatars/58a9abce945e71a65abc8a54085de6d7.svg","isPro":false,"fullname":"oh sehun","user":"sehun","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.06741.md"}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2606.06741 in a model README.md to link it from this page.

Cite arxiv.org/abs/2606.06741 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2606.06741 in a Space README.md to link it from this page.