[{"data":1,"prerenderedAt":147},["ShallowReactive",2],{"project-48579116":3},{"id":4,"name":5,"line":6,"line_source":7,"title":8,"url":9,"domain":10,"maker":11,"posted_at":12,"year":13,"points":14,"comments":15,"rank_in_year":16,"shown_in_year":17,"top":18,"tier":19,"category":20,"subcategory":21,"format":22,"keywords":23,"reception":29,"sentiment":30,"stance_mix":31,"scored_comments":35,"aspects":36,"strip":39,"post_text":40,"what":41,"said":55,"maker_in_thread":83,"shape":94,"grid":98,"attempts":110,"same_week":111,"maker_shows":14,"maker_hidden":140,"reader_version":141,"card":142},"48579116","LegalRabbit DOCX plugin","Cowork\u002FCodex DOCX plugin. Uses 2x fewer tokens than the docx skill","title","Show HN: Cowork\u002FCodex DOCX plugin. Uses 2x fewer tokens than the docx skill","https:\u002F\u002Fgithub.com\u002FLegalRabbit-AI\u002Flegalrabbit-docx-claude-plugin","github.com","tanin",1781743776,2026,5,6,3727,4366,86,"shown","productivity","document editing plugin","other",[24,25,26,27,28],"docx","html conversion","tracked changes","redlining","ai plugin","quiet",null,{"positive":32,"neutral":33,"mixed":34,"critical":34},2,1,0,3,{"execution":37,"design":38},{"positive":34,"critical":34},{"positive":33,"critical":34},[35,34,14,34,14,34],"Hi HNers,\n\nI'd like to share our DOCX plugin for Cowork and Codex.\n\nIt uses 2-5x fewer tokens compared to the traditional docx skill because it doesn't write any code nor execute python\u002Fnode script. It is also much more reliable.\n\nOur DOCX plugin converts docx\u003C->html bidirectionally. This means AI only operates on HTML. AI is excellent and very efficient when it comes to HTML.\n\nMost libraries (if not all) support docx->html, but none supports html->docx. This is what is novel about our approach.\n\nHere's the demo: https:\u002F\u002Fdrive.google.com\u002Ffile\u002Fd\u002F1UNlUJYwkNX3NiANDkLLb3UoRSms2dXCU\u002Fview?usp=drive_",{"problem":42,"audience":30,"pricing":30,"motivation":48,"time_to_build":30,"stack":52,"openness":53,"stage":54,"team_size":53},{"value":43,"source":44,"evidence":45},"AI editing of DOCX files uses many tokens and is unreliable when it writes code","maker",[46],{"id":47},"post",{"value":49,"source":44,"evidence":50},"Used in-house for redlining legal documents",[51],{"id":47},[],"unknown","launched",{"praise":56,"objections":72,"requests":73,"questions":74,"alternatives":82},[57,65],{"value":58,"source":59,"evidence":60},"Approach to converting under the hood is appreciated","commenter",[61],{"id":62,"by":63,"text":64},"48580263","dev-kdrainc","Thanks for sharing! I like your approach to working under the hood! Great job",{"value":66,"source":59,"evidence":67},"Commenter plans to try it",[68],{"id":69,"by":70,"text":71},"48582544","StahlGuo","I would try it today, sounds good",[],[],[75],{"value":76,"source":59,"evidence":77},"Whether tracked changes and comments are preserved in the HTML layer",[78],{"id":79,"by":80,"text":81},"48579476","xms17189","Interesting approach. Does keeping the model in HTML also preserve enough structure for tracked changes\u002Fcomments, or do you handle those as a separate layer when converting back…",[],{"comments":35,"answered":33,"key":84},[85,88,91],{"id":86,"text":87},"48579520","Thank you! My thesis is that an intermediate layer would eventually end up being equivalent to the docx format, so I've decided not to have any intermediate representation. We…",{"id":89,"text":90},"48591919","Please let me know if you have questions or suggestions.",{"id":92,"text":93},"48593598","Thank you!",{"commenters":95,"max_depth":32,"maker_replies":35,"first_comment_mins":96,"lifespan_hours":97},4,48,24,[99,101,103,105,107,109],[79,34,35,100,80,34],"project\u002Fneutral\u002Fexecution",[86,34,34,102,11,33],"project\u002Fneutral\u002Fdesign",[69,33,14,104,70,34],"project\u002Fpositive\u002Fnone",[89,33,34,106,11,33],"maker\u002Fneutral\u002Fnone",[62,32,14,108,63,34],"maker\u002Fpositive\u002Fdesign",[92,32,34,106,11,33],[],[112,117,122,127,132,136],{"id":113,"name":114,"points":115,"category":116},"48591348","in the Weights",471,"ai",{"id":118,"name":119,"points":120,"category":121},"48645437","TikZ Editor",451,"dev_tools",{"id":123,"name":124,"points":125,"category":126},"48616304","Chess-Inspired Roguelike",433,"games",{"id":128,"name":129,"points":130,"category":131},"48608570","TownSquare",278,"web_saas",{"id":133,"name":134,"points":135,"category":121},"48660267","Nub",277,{"id":137,"name":138,"points":139,"category":126},"48629213","puzzlelair.com",251,false,"v2",{"id":143,"name":5,"line":6,"ai":140,"cat":20,"year":13,"points":14,"comments":15,"rec":29,"s":30,"top":18,"kw":144,"maker":11,"domain":10,"tier":19,"attempts":34,"strip":145,"url":9,"mix":146,"praise":58,"objection":-1,"sub":21},48579116,[24,25,26,27,28],"305050",[32,33,34,34],1793991157154]