[{"data":1,"prerenderedAt":313},["ShallowReactive",2],{"maker-eigenvalue":3},[4,34,57,79,101,123,144,167,185,205,222,237,254,271,292],{"id":5,"name":6,"line":7,"ai":8,"cat":9,"year":10,"points":11,"comments":12,"rec":13,"s":14,"top":15,"kw":16,"maker":22,"domain":23,"tier":24,"attempts":25,"strip":26,"url":27,"mix":28,"praise":31,"objection":32,"sub":33},36976333,"llama2_aided_tesseract","Using LLama2 to Correct OCR Errors",false,"ai",2023,54,10,"curious",-0.17,34,[17,18,19,20,21],"llama2","ocr correction","tesseract","llm","text recognition","eigenvalue","github.com","notable",0,"3030303100","https:\u002F\u002Fgithub.com\u002FDicklesworthstone\u002Fllama2_aided_tesseract",[25,29,25,30],4,1,"Idea of LLM-based proofreading of OCR is promising","Llama2 hallucinates new sentences, needing filtering","ocr error correction",{"id":35,"name":36,"line":36,"ai":8,"cat":9,"year":10,"points":37,"comments":38,"rec":39,"s":40,"top":41,"kw":42,"maker":22,"domain":23,"tier":24,"attempts":48,"strip":49,"url":50,"mix":51,"praise":54,"objection":55,"sub":56},37133163,"Llama2 Embeddings FastAPI Server",178,31,"roasted",-0.34,11,[43,44,45,46,47],"llama2 embeddings","embeddings api","semantic search","self-hosted embeddings","fastapi",2,"10103001313000001310503030110030","https:\u002F\u002Fgithub.com\u002FDicklesworthstone\u002Fllama_embeddings_fastapi_service",[30,52,25,53],7,8,"Clean and simple codebase","Raw Llama2 likely underperforms dedicated embedding models like bge, gte and e5","embeddings server",{"id":58,"name":59,"line":60,"ai":8,"cat":61,"year":10,"points":62,"comments":63,"rec":13,"s":64,"top":65,"kw":66,"maker":22,"domain":23,"tier":24,"attempts":25,"strip":72,"url":73,"mix":74,"praise":76,"objection":77,"sub":78},37234887,"Fast vector similarity","Fast vector similarity using Rust and Python","data",141,40,-0.05,15,[67,68,69,70,71],"vector similarity","hoeffding's d","rust python","bootstrap similarity","embedding comparison","1033030050300500300030000103001000000000","https:\u002F\u002Fgithub.com\u002FDicklesworthstone\u002Ffast_vector_similarity",[48,52,25,75],3,"Fast performance versus NumPy","Library can consume all CPU cores and compete with other workloads","similarity measures library",{"id":80,"name":81,"line":82,"ai":8,"cat":83,"year":10,"points":84,"comments":85,"rec":39,"s":86,"top":87,"kw":88,"maker":22,"domain":23,"tier":94,"attempts":25,"strip":95,"url":96,"mix":97,"praise":98,"objection":99,"sub":100},37600019,"automatic_log_collector_and_analyzer","My Single-File Python Script I Used to Replace Splunk in My Startup","infrastructure",313,79,-0.06,5,[89,90,91,92,93],"log analysis","log collection","splunk alternative","sqlite logs","datasette","legend","0050000003113511000000000000051033015031","https:\u002F\u002Fgithub.com\u002FDicklesworthstone\u002Fautomatic_log_collector_and_analyzer",[52,48,30,12],"Simple, practical tool that suits small-scale needs","Does not scale to large enterprise log volumes","log collection and analysis",{"id":102,"name":103,"line":104,"ai":8,"cat":9,"year":10,"points":105,"comments":106,"rec":107,"s":108,"top":109,"kw":110,"maker":22,"domain":23,"tier":24,"attempts":25,"strip":116,"url":117,"mix":118,"praise":120,"objection":121,"sub":122},38236198,"bulk_transcribe_youtube_videos_from_playlist","Bulk Creation of Transcripts from YouTube Playlists with Whisper",125,43,"loved",0.26,17,[111,112,113,114,115],"whisper transcription","youtube transcript","bulk transcription","speech to text","faster-whisper","3003100000300555050003000330035500000000","https:\u002F\u002Fgithub.com\u002FDicklesworthstone\u002Fbulk_transcribe_youtube_videos_from_playlist",[119,52,25,30],6,"Whisper transcripts are more accurate than YouTube's auto-generated ones.","Speaker diarization is not solved in this tool.","speech transcription",{"id":124,"name":125,"line":126,"ai":8,"cat":127,"year":10,"points":85,"comments":128,"rec":129,"s":130,"top":131,"kw":132,"maker":22,"domain":23,"tier":24,"attempts":48,"strip":138,"url":139,"mix":140,"praise":141,"objection":142,"sub":143},38249730,"sqlalchemy_data_model_visualizer","Easily Visualize Your SQLAlchemy Data Models in a Nice SVG Diagram","dev_tools",18,"mixed",0.18,26,[133,134,135,136,137],"sqlalchemy","erd diagram","graphviz","data model visualization","svg","030500050135050310","https:\u002F\u002Fgithub.com\u002FDicklesworthstone\u002Fsqlalchemy_data_model_visualizer",[29,75,25,48],"Useful for onboarding new devs and CI schema diagrams","Fails with older SQLAlchemy versions due to declarative_base import","diagram generator",{"id":145,"name":146,"line":147,"ai":8,"cat":148,"year":149,"points":150,"comments":85,"rec":39,"s":151,"top":152,"kw":153,"maker":22,"domain":158,"tier":24,"attempts":25,"strip":159,"url":160,"mix":161,"praise":164,"objection":165,"sub":166},39142290,"Alzheimer's Buddy","Use Flashing Light and Sound at 40Hz","hardware",2024,96,-0.37,25,[154,155,156,157],"40hz stimulation","alzheimer's","flicker light","gamma entrainment","alzheimersbuddy.com","1303011111101111111331310335333313035311","https:\u002F\u002Falzheimersbuddy.com\u002F",[48,162,25,163],21,23,"Interesting concept and project","Flashing damaged monitors, causing lasting flicker","light and sound stimulation",{"id":168,"name":169,"line":170,"ai":8,"cat":171,"year":149,"points":172,"comments":173,"rec":39,"s":40,"top":128,"kw":174,"maker":22,"domain":23,"tier":24,"attempts":25,"strip":179,"url":180,"mix":181,"praise":182,"objection":183,"sub":184},41162505,"visual_astar_python","Visual A* pathfinding and maze generation in Python","learning_science",135,38,[175,176,177,178],"a* pathfinding","maze generation","python visualization","matplotlib animation","01000111000501000030500101110031030100","https:\u002F\u002Fgithub.com\u002FDicklesworthstone\u002Fvisual_astar_python",[48,75,25,41],"Interesting idea for applying A* heuristics to real problems","Code and README appear largely LLM-generated, with repetitive style and no OOP structure","pathfinding visualization",{"id":186,"name":187,"line":188,"ai":8,"cat":189,"year":149,"points":190,"comments":191,"rec":13,"s":192,"top":75,"kw":193,"maker":22,"domain":23,"tier":94,"attempts":25,"strip":198,"url":199,"mix":200,"praise":202,"objection":203,"sub":204},41203306,"LLM-aided OCR","Correcting Tesseract OCR errors with LLMs","productivity",479,172,-0.2,[194,19,195,196,197],"ocr","llm correction","pdf to markdown","scanned books","3110113105000413300033003110325132031331","https:\u002F\u002Fgithub.com\u002FDicklesworthstone\u002Fllm_aided_ocr",[119,201,48,128],29,"Multi-stage prompting with cheap fast models works well","LLM can silently alter numbers and names, risky for critical data","ocr document conversion",{"id":206,"name":207,"line":207,"ai":8,"cat":9,"year":149,"points":41,"comments":29,"rec":208,"s":209,"top":210,"kw":211,"maker":22,"domain":23,"tier":215,"attempts":25,"strip":216,"url":217,"mix":218,"praise":219,"objection":220,"sub":221},41224623,"LLM Aided Transcription Improvement","quiet",null,71,[212,213,20,114,214],"transcription","whisper","markdown","shown","3030","https:\u002F\u002Fgithub.com\u002FDicklesworthstone\u002Fllm_aided_transcription_improvement",[25,48,25,25],"Useful for turning voice memos and rambling recordings into readable text","Garbage in, garbage out: poor Whisper output on noisy audio limits results","transcript cleanup",{"id":223,"name":224,"line":225,"ai":8,"cat":189,"year":149,"points":75,"comments":87,"rec":208,"s":209,"top":226,"kw":227,"maker":22,"domain":230,"tier":215,"attempts":48,"strip":231,"url":232,"mix":233,"praise":234,"objection":235,"sub":236},41636074,"YouTube Transcript Optimizer","Turn Videos into Polished Documents",94,[112,228,229,212],"video to document","quiz generator","youtubetranscriptoptimizer.com","30300","https:\u002F\u002Fyoutubetranscriptoptimizer.com",[25,30,30,25],"Good tool","YouTube may block the service at scale","content repurposing",{"id":238,"name":239,"line":240,"ai":8,"cat":9,"year":241,"points":242,"comments":52,"rec":208,"s":209,"top":243,"kw":244,"maker":22,"domain":23,"tier":215,"attempts":25,"strip":249,"url":250,"mix":251,"praise":252,"objection":253,"sub":245},43364640,"LLM-docs","LLM-docs, software documentation intended for consumption by LLMs",2025,28,45,[245,246,247,248],"llm documentation","rag","documentation distillation","llm context","0303003","https:\u002F\u002Fgithub.com\u002FDicklesworthstone\u002Fllm-docs",[25,48,30,25],"Idea of machine-readable docs is valuable and fertile ground","Distilled docs may not be what LLMs need; retrieval is the harder problem",{"id":255,"name":256,"line":257,"ai":8,"cat":127,"year":241,"points":258,"comments":48,"rec":208,"s":209,"top":259,"kw":260,"maker":22,"domain":23,"tier":215,"attempts":48,"strip":265,"url":266,"mix":267,"praise":268,"objection":269,"sub":270},45793684,"mcp_agent_mail","Give your coding agents the ability to message each other",14,63,[261,262,263,264],"coding agents","agent messaging","mcp","multi-agent","30","https:\u002F\u002Fgithub.com\u002FDicklesworthstone\u002Fmcp_agent_mail",[25,25,30,25],"Works well and is quick and easy to set up.","Replicating corporate processes with agents may be broken.","agent coordination",{"id":272,"name":273,"line":273,"ai":8,"cat":171,"year":274,"points":275,"comments":276,"rec":129,"s":277,"top":278,"kw":279,"maker":22,"domain":285,"tier":24,"attempts":25,"strip":286,"url":287,"mix":288,"praise":289,"objection":290,"sub":291},46833754,"Phage Explorer",2026,127,35,-0.07,12,[280,281,282,283,284],"phage","bacteriophage","genome visualization","virology education","protein structure","phage-explorer.org","31001100000300000000133000010030555","https:\u002F\u002Fphage-explorer.org\u002F",[75,30,29,87],"Good premise and interesting educational goal","Scientific inaccuracies in diagrams and genetic representations","interactive biology education",{"id":293,"name":294,"line":295,"ai":8,"cat":127,"year":274,"points":296,"comments":12,"rec":297,"s":298,"top":299,"kw":300,"maker":22,"domain":306,"tier":215,"attempts":25,"strip":307,"url":308,"mix":309,"praise":310,"objection":311,"sub":312},46986644,"FrankenTUI","FrankenTUI in the Browser",22,"debated",0.19,44,[301,302,303,304,305],"terminal ui","tui framework","xterm alternative","browser terminal","react terminal widget","frankentui.com","5350015101","https:\u002F\u002Ffrankentui.com\u002Fweb",[75,30,25,75],"Impressive demo that looks like it works","Broken interactions: text selection, scrollbar, and mouse clicks misbehave","terminal ui framework",1793991280087]