We also provide live demo at: <a href=\"https://designanything.ai/\" rel=\"nofollow\">https://designanything.ai/</a>, though, we recommend to locally install for the best experience.<br>We also welcome the community to submit issues or propose PR, together, we can continously improve autodesign.</p>\n","updatedAt":"2026-08-14T02:53:02.406Z","author":{"_id":"653cb809b424289c5f384a02","avatarUrl":"/avatars/a1565ab5ae51075c75d6857d64c426a8.svg","fullname":"YaxinLuo","name":"YaxinLuo","type":"user","isPro":true,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":1,"isUserFollowing":false}},"numEdits":2,"identifiedLanguage":{"language":"en","probability":0.8667986989021301},"editors":["YaxinLuo"],"editorAvatarUrls":["/avatars/a1565ab5ae51075c75d6857d64c426a8.svg"],"reactions":[],"isReport":false}},{"id":"6a7ecc28f8cefdde6efc083d","author":{"_id":"653cb809b424289c5f384a02","avatarUrl":"/avatars/a1565ab5ae51075c75d6857d64c426a8.svg","fullname":"YaxinLuo","name":"YaxinLuo","type":"user","isPro":true,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":1,"isUserFollowing":false},"createdAt":"2026-08-14T08:04:56.000Z","type":"comment","data":{"edited":false,"hidden":false,"latest":{"raw":"AutoDesign For AutoDesign Case：\n\n\n","html":"<p>AutoDesign For AutoDesign Case：</p>\n<p><a href=\"https://cdn-uploads.huggingface.co/production/uploads/653cb809b424289c5f384a02/G4NHCEZj2dDG3TX5L8FHw.png\" rel=\"nofollow\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/653cb809b424289c5f384a02/G4NHCEZj2dDG3TX5L8FHw.png\" alt=\"截屏2026-08-14 14.52.58\"></a></p>\n","updatedAt":"2026-08-14T08:04:56.464Z","author":{"_id":"653cb809b424289c5f384a02","avatarUrl":"/avatars/a1565ab5ae51075c75d6857d64c426a8.svg","fullname":"YaxinLuo","name":"YaxinLuo","type":"user","isPro":true,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":1,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.3111075758934021},"editors":["YaxinLuo"],"editorAvatarUrls":["/avatars/a1565ab5ae51075c75d6857d64c426a8.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2608.13560","authors":[{"_id":"6a7e76e342823931a1f1760d","name":"Yaxin Luo","hidden":false},{"_id":"6a7e76e342823931a1f1760e","name":"Haobin Jiang","hidden":false},{"_id":"6a7e76e342823931a1f1760f","name":"Jialv Zou","hidden":false},{"_id":"6a7e76e342823931a1f17610","name":"Xu Huang","hidden":false},{"_id":"6a7e76e342823931a1f17611","name":"Wenhao Yan","hidden":false},{"_id":"6a7e76e342823931a1f17612","name":"Haodong Li","hidden":false},{"_id":"6a7e76e342823931a1f17613","name":"Zhengrong Yue","hidden":false},{"_id":"6a7e76e342823931a1f17614","name":"Jing Li","hidden":false},{"_id":"6a7e76e342823931a1f17615","name":"Xiaofu Chen","hidden":false},{"_id":"6a7e76e342823931a1f17616","name":"Xiaohan Zhao","hidden":false},{"_id":"6a7e76e342823931a1f17617","name":"Jiacheng Liu","hidden":false},{"_id":"6a7e76e342823931a1f17618","name":"Jiacheng Cui","hidden":false},{"_id":"6a7e76e342823931a1f17619","name":"Zhiqiang Shen","hidden":false},{"_id":"6a7e76e342823931a1f1761a","name":"Xiaotong Li","hidden":false}],"publishedAt":"2026-08-13T00:00:00.000Z","submittedOnDailyAt":"2026-08-14T00:00:00.000Z","title":"AutoDesign: Meta-Harness Optimization for Long-Horizon Agentic Design","submittedOnDailyBy":{"_id":"653cb809b424289c5f384a02","avatarUrl":"/avatars/a1565ab5ae51075c75d6857d64c426a8.svg","isPro":true,"fullname":"YaxinLuo","user":"YaxinLuo","type":"user","name":"YaxinLuo"},"summary":"Transforming multimodal sources into condensed and structured media outputs can be fundamentally conceptualized as a long-horizon agentic process centered on a model-harness system. While an ideal harness system should align with human design priors and accumulate reusable experience through empirical exploration to drive recursive self-improvement, existing paradigms remain static and fall short of this capability. In this paper, we present AutoDesign, a framework that aligns with human design priors, where a meta-harness optimizer guides a code agent to recursively improve harness based on rollout feedback. To instantiate and evaluate this framework, we focus on the academic paper-to-poster generation task and introduce PosterBench, comprising a 100-paper Main Track spanning five disciplines and PosterBench-mini, a shared 10-paper subset for controlled evaluation. On the PosterBench Main Track, AutoDesign achieves the highest score of 78.32, surpassing the closed-source commercial system Claude Design by 7.45 points. Across seven controlled code-agent-model configurations, integrating the learned DesignHarness consistently improves performance, increasing the average PosterBench Score from 54.99 to 67.39 (+12.4%). In a fully autonomous long-horizon loop, it executes 253 tool calls and 11 editing turns within 40 minutes for under $3, reaching average conference-poster quality in human evaluation. A system-blind human study further demonstrates that AutoDesign achieves the highest human preference among evaluated systems.","upvotes":28,"discussionId":"6a7e76e442823931a1f1761b","projectPage":"https://autodesign.designanything.ai/","githubRepo":"https://github.com/Yaxin9Luo/AutoDesign","githubRepoAddedBy":"user","ai_summary":"AutoDesign uses a meta-harness optimizer to recursively improve a code agent for structured media generation, achieving state-of-the-art results on paper-to-poster synthesis.","ai_keywords":["meta-harness optimizer","code agent","DesignHarness","rollout feedback","long-horizon agentic process"],"ai_summary_model":"thinkingmachines/Inkling-Small","githubStars":7,"organization":{"_id":"6282108b1c4fdf630c7943a4","name":"meituan","fullname":"meituan","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/61ac8f8a00d01045fca0ad2f/4TJKRXMsRyZsyi4H3rWsh.jpeg"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"67b9732487172149b94dcd46","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/5QR9mxfReWrcwiOE6u7xF.png","isPro":false,"fullname":"Xiaohan Zhao","user":"ritianzhaoxx","type":"user"},{"_id":"6425dac02a423daad529fec1","avatarUrl":"/avatars/b796538366d7fabc4a4aa27549370260.svg","isPro":false,"fullname":"Jiacheng Liu","user":"zcahjl3","type":"user"},{"_id":"66f54fb4309542d43f5c7d3e","avatarUrl":"/avatars/75c74b0dca38c99f9c1df34d3601e826.svg","isPro":false,"fullname":"wenhao","user":"wenhaoyan77","type":"user"},{"_id":"620783f24e28382272337ba4","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/620783f24e28382272337ba4/zkUveQPNiDfYjgGhuFErj.jpeg","isPro":false,"fullname":"GuoLiangTang","user":"Tommy930","type":"user"},{"_id":"65ddea8b2d26e59a5a33330f","avatarUrl":"/avatars/3104ddafd6dda3c05ea9a771dbf2deeb.svg","isPro":false,"fullname":"li haodong","user":"mickyhimself","type":"user"},{"_id":"6523a8e98fc48849920451bc","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6523a8e98fc48849920451bc/4ASm-gyh72APgQwcnbIaE.jpeg","isPro":false,"fullname":"Wonderful","user":"wonderful9462","type":"user"},{"_id":"67543820c3af453d7b3e1d5e","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/67543820c3af453d7b3e1d5e/RbAZ9AQlxpy5E5Is-QN8b.jpeg","isPro":false,"fullname":"Dingming Li","user":"lidingm","type":"user"},{"_id":"645f719aa3e00932c9ef2dc8","avatarUrl":"/avatars/a6436453cfccf0351ac83719051db415.svg","isPro":false,"fullname":"Doctor-James","user":"Doctor-James","type":"user"},{"_id":"65fe6006fc78eace53109b7a","avatarUrl":"/avatars/55a896b89bcf5e390195b442cf5076be.svg","isPro":false,"fullname":"LI, XIAOTONG","user":"XiaotongLi97","type":"user"},{"_id":"675d78ecb82471cebbf5c855","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/xsm99BUEUxYpYWa5Qrv54.png","isPro":false,"fullname":"daisiuki","user":"daisiuki","type":"user"},{"_id":"68106c88b924dd6c328889c2","avatarUrl":"/avatars/8accf835b711bffa2ea307158950ab33.svg","isPro":false,"fullname":"Hongbo Peng","user":"M1chaelPeng","type":"user"},{"_id":"661f9a1b142a51d630dc165d","avatarUrl":"/avatars/b0d5325c4ce7cb632bc22766ebbecdec.svg","isPro":false,"fullname":"whalelin_Lin","user":"whalelin","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"6282108b1c4fdf630c7943a4","name":"meituan","fullname":"meituan","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/61ac8f8a00d01045fca0ad2f/4TJKRXMsRyZsyi4H3rWsh.jpeg"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2608/2608.13560.md","query":{}}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2608.13560 in a model README.md to link it from this page.

Cite arxiv.org/abs/2608.13560 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2608.13560 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.