🔥 話題の対戦

🔥 nekohasekai, you're an "OSS workhorse" whose personal repos are nothing but empty shells, riding entirely on SagerNet org's coattails. Meanwhile rprx, despite having barely any solo projects, has welded ML-KEM-768 into Xray-core — a real nuclear-grade contribution that crushes your ecosystem impact. All your PRs just make you a high-level mercenary, not a leader.

🔥 skiyee, with 16 followers you dare to submit trivial PRs to a 18k-star repo like a janitor? Three CNAME-change PRs treating js.org as a punch card, and among your 12 original projects only one organization-owned 165-star fig leaf, the rest are 0–9 star e-waste. That’s not open source—it’s digital panhandling. Your ecosystem impact is a full tier below xiaohe0601; he actually shoulders a 556-star nutui repo while your tag is literally “Personal Project Wasteland.”

🔥 It's a foregone conclusion that hikariming crushes lvy010 with a massive 24.8-point gap. Among the six dimensions, lvy only scraped a full score in community influence by stacking followers, while all core dimensions including original project quality, contribution quality and ecosystem impact are completely blown out. You have 176 forked repos out of 188 warehouses as hoarded junk, your so-called "hardest" 1275-star project is just a patchwork of your four years of college notes, and all your e

🔥 dreamlike-ocean crushes lvy010 outright on core metrics of contribution quality and ecosystem impact, backed by 32 hardcore PRs to Netty, the 35k-star top-tier project; lvy’s 188 repos with 176 forked, and his most hyped 1275-star repo is just a stitched-together product of 4 years of college notes, with laughably poor original project quality, and his community influence score piled up by fan filters and note repos is completely worthless against tangible open source contributions, the 21.5-poi

🔥 B edges out A by a meager 1.1 points in a dead-even NPC-tier matchup that’s basically two underperformers pecking at each other. A only has one presentable project: a 2253-star Jupyter Notebook, while all his other original repos are sub-100 star filler, half his PRs get closed by maintainers, and he’s clearly using his Shanghai AI Lab affiliation as an open source resume prop, with 133 followers so low-quality they could practically support fish. B is only propped up by a single ICML Oral proje

🔥 Wow, this 16.7-point blowout is absolutely brutal: A's original project quality is so pathetic that his top project only has 127 stars, with nearly 40% of his contributions being low-effort edits to docs, YAML files and robots.txt for zero-star repos, getting totally crushed by B in both ecosystem impact and community influence; B has the Linux kernel as a top-tier original project, with full ecosystem impact and solid contribution quality, those "high-star contributions" you scraped together by

🔥 Yo, A, you wear the titles of "OSS Workhorse" and "PR Doer" pretty proudly, don’t you? You crush B in four dimensions: original project quality, contribution quality, community influence and activity authenticity, yet you scored a flat 0 in ecosystem impact? All that grinding you do is completely invisible to the open source ecosystem, not even a peep? Look at B, your original project quality and contribution quality are so trash they’re barely passable, but you scraped a narrow 0.7 point win ju

🔥 syuan03 ekes out a 3.7-point victory over liuziyu77, and this even-tier result is basically a tie. liuziyu77 dominates in original project quality, ecosystem impact and community influence, but has trash contribution quality with half their PRs closed by maintainers and contribution water content high enough to raise fish, relying entirely on one Jupyter Notebook to carry their stats, and their 133 followers are just the result of using GitHub as an open source resume decoration; syuan03 wins on

🔥 jizizr, five years on GitHub and you've got barely 30 stars—your only "valuable" contribution is fixing a typo in a 7.9k-star repo, and out of 5 merged PRs, two are trivial installer script tweaks. All 31 forks are gathering dust. Your open-source resume is too pathetic to even decorate a CV. Meanwhile mogumc may be a free grinder for big projects, but they've racked up 8 PRs and 12 commits, crushing you in ecosystem impact, contribution quality, and original projects. The only dimension you "wo

🔥 This isn't even a dev showdown, it's a one-sided total crush with zero suspense. nihildigit only scores 11.5/18 on original project quality, his most popular personal repo has just 48 stars, and all his credibility comes from contributing core engineering code to other people's 10k+ star corporate projects — he's basically a free senior outsourcer of the open source world. His community influence score is a laughable 2.6/8, his 7-year old account has fewer followers than a tenth of the repos he

🔥 Both are GOD-tier Android OSS veterans, but yujincheng08 only edges out vvb2060 by 2.3 points in an even matchup, and neither has any right to act high and mighty. vvb2060, you’ve had an account for 9 years and 4760 followers watching you, yet you’ve spent all your energy pumping out 258 commits for LSPosed and 181 PRs for Magisk like a free engine for big projects, and your highest-starred original project barely hits 2k? You can’t even pull out a decent hit project of your own, and your ecosys

🔥 This 2.8-point narrow win is 100% deserved, as B’s original project quality and contribution quality get straight-up crushed by A: outside of the repo that leeches JD’s NutUI org name to hit 556 stars, all his personal original projects top out at just 165 stars, nearly 40% of his external PRs are useless doc-editing and window-wiping fluff, and he even spammed two button variant PRs to wot-ui to pad his contribution count. Both original quality and contribution quality are garbage, his communit

🔥 lvy010, your only won dimension is "community influence" – thanks to 3,640 simping followers and 47 followings? That's a "fan-filter clique" victory with zero substance. Meanwhile kohakublueleaf: LyCORIS 2.5k stars, 112 PRs to ComfyUI, 2% rejection rate, crushing you on every real metric. Your flagship project is a glorified college note dump, 188 forked repos are just digital hoarding, and you close your own PRs – that's not contributing, that's performance art.

🔥 What a narrow 9.2-point edge win for luochen211, he almost lost his pants: he dragged the total score up with a maxed 20 ecosystem impact score and 21.9 contribution quality, but 67% of his merged PRs are just doc/config tweaks for a single 59k-star big tech repo, his 23 personal repos top out at only 2 stars, and he even writes "Focus vibe coding" in his bio? Your vibe is just being a hype man for other people's projects, huh? As for lvy010, he has advantages in original project quality and com

🔥 makinoharashoko edges out mefengl by 7 points in a narrow win, and the result comes down to raw dimension stats: you mefengl, tagged as "OSS Workhorse" and "Follow Maniac", have trash original project quality — your highest-star personal repo is a 134-star VSCode plugin literally named "too lazy to write commit messages", your total personal stars are only 416, all from ghosting PRs to 10k-star repos like tinygrad and taro. You follow 305 people but only have 133 followers? You're a textbook ope

🔥 dramforever wins by a razor-thin 3.2-point margin over nekohasekai, a textbook even-bucket nail-biter. A dominates across three key dimensions: original project quality, contribution quality and ecosystem impact, but is the classic "punches hard for others, leaves his own yard barren" archetype: 440 followers, 0 following, 44 commits to the Linux kernel, 64 merged PRs to NixOS, 57% of his PRs go to the 56-star openRuyi as free labor, and his most popular personal project only hits 224 stars – al

🔥 leaderonepro edges out lvy010 by only 10.5 points, and this win is nothing to show off: your original project quality is trash, scoring only 8.3 out of 18, you scrape by on contribution quality, ecosystem impact and activity authenticity by spamming filler PRs that add LLM configs and rewrite READMEs for 10k-star big repos like CherryHQ and sipeed, you only have 62 followers which is less than one tenth of your opponent’s, and that “open source grunt” and “doc decorator” persona is just full of

🔥 konakona418 edges out lvy010 purely on the back of hard-earned ecosystem impact and contribution quality from fixing bugs in mainstream projects like raylib and libuv. The former's tangible contributions to big-name repos straight up expose the latter, who can't even land a core PR, relying entirely on forked note repos and fan bias to prop up their facade. Even with your higher original project score and 3k+ followers, your open source contribution credibility is way lower than his, even if he'

🔥 shadcn, you scored a big fat zero across all six sub-metrics and still scraped an edge win? Guess GitHub's scoring system just hands out free points to big names? Don't get cocky - if we judged by actual hard skills you wouldn't even be qualified to tie hyaciovo's shoelaces. Your tag says "zero-follower big shot" and you're padding your score with nothing but fame, this edge win is basically a system bug, and you have the audacity to call yourself a GOD? And hyaciovo, stop whining. You won every

🔥 This 0.3-point razor-thin win is basically etching the awkwardness of being an open source laborer on everyone's face: B, your original project quality and contribution quality scores are indeed half a step higher than A's, but you get absolutely crushed in ecosystem impact, community influence, and activity authenticity. Your biggest repo only has 30 stars, and even after merging 341 PRs for Stockfish, you still can't get any traction on your own profile? You're a straight-up open source temp w

🔥 That 0.1-point razor-thin win is basically held up by contribution quality and ecosystem impact, huh? Sure, ilyfairy's original projects are garbage — 113 repos over 7 years with barely 123 total stars, his highest-starred one is just a Don't Starve server query tool, but every PR he slaps on popular repos is a core feature, so his ecosystem impact is through the roof. Meanwhile slimenull's best original repo only has 198 stars, nearly half his external contributions are just doc edits, site twe

🔥 What a farce, two of the biggest legends of the Vue ecosystem ended up in a dead-heat with only a 3.1 point gap, no less. yyx990803, you’re sitting on 400k+ total stars from Vue and Vite, but your original project quality score is 0.7 points lower than yunyoujun’s, and your activity authenticity is a full 2.4 points behind — so that 40% of your recent PRs that just change logos and tweak doc themes are just you grinding for GitHub activity KPIs? yunyoujun’s Valaxy only has 1115 stars, but he bea

🔥 Both are GOD-tier, but the 3.6 point gap is barely enough for glavo to edge out the win. torvalds maxes out ecosystem impact and community influence, but all his GitHub contributions go to dive software and guitar mods, getting completely crushed by glavo in contribution quality and activity authenticity — he's like an open source god who shoves his Linux kernel crown project in the corner to mess around with toys. On the flip side, glavo gets碾'd to dust by torvalds in ecosystem impact and commu

🔥 glavo's crushing victory is well-earned, but his open-source tough guy persona is basically held up by a single 9.5k-star HMCL repo: 89 of his 115 original repos are empty forks with zero substance, his personal original projects have less than 800 total stars with a high of only 108, he even submitted 21 low-effort doc-editing PRs to OpenJDK to leech top-tier project clout, and nearly a third of his contributions are trivial doc polishing and site window cleaning grunt work, so his "OSS workhor

🔥 hellock flexes 11 years of OSS workhorse pedigree with near-perfect contribution quality (25.8/27) and maxed ecosystem impact, while liuziyu77’s sole win in original project quality is carried by a single 2253-star notebook—everything else? Bottom-tier: 7.3/27 contribution quality, one PR closed on arrival, activity so hollow you could farm tilapia. This crush is a reality check between a real core contributor and a resume-padding cosplayer.

🔥 This is a total public curb stomp, 001promax gets completely crushed by xream across all six sub-metrics and public actions: your original project quality of 8.9/18 is less than half of your opponent's, your ecosystem impact of 6.6/20 gets absolutely steamrolled by 19.1 points, 4 years on GitHub only netted you 34 stars which is less than a drop in xream's bucket, you've got 8 self-closed external PRs and 4 original projects that peak at 29 stars, you're just a pure GitHub resume padding loser,

🔥 xream snags a 4.2-point edge win, all carried by hard metrics: it absolutely crushes nekohasekai on original project quality and contribution quality. While nekohasekai maxes out ecosystem impact, community influence and activity authenticity, it falls nearly 5 points short on original project quality and 0.5 points on contribution quality, only keeping the gap narrow thanks to its three perfect scores. For xream, it hits a full 18/18 on original project quality backed by its 1863 solo commits t

🔥 That 0.9 gap pins both GOD-tier devs right on the shame wall of an even match? rprx, your XTLS core contributor cred is rock solid — you crush xream on contribution quality, ecosystem impact and community influence, but 1050 days of radio silence plus your highest original repo sitting at 4 stars? The "zero-doc spammer" tag is basically tattooed on your forehead. xream, your open-source workaholic persona is well-built with top-tier activity authenticity, but your contribution quality gets destr

🔥 While A only trails by 1.4 points overall, he leads in account maturity, original project quality and contribution quality, but gets hard crushed by B in ecosystem impact, community influence and activity authenticity. A’s 17 original repos have a total of only 17 stars, and nearly 40% of his PRs are just filler work like dumping materials into heu-icicles or fixing spreadsheets, making him a pure GitHub resume decorator. B wins on ecosystem and activity metrics, but he’s a 5-month-old newbie wi

🔥 Two NPC-tier rookies pecking at each other with a 0.7-point gap, even match so no one gets to mock the other: A, you're the king of open-source self-indulgence, 6 of your 7 PRs are filler in your own 0-1 star repos. Your original project quality is better than B's, but your ecosystem impact and community influence are completely crushed by B, your total stars are fewer than pearls in milk tea, your 4 fans aren't even enough for a mahjong table, and your open-source persona packaging is thicker t

🔥 These two NPC-tier devs only have a 2.6-point gap yet got lumped into the "even" bucket, what a bunch of noobs pecking at each other. B relies on its ICML Oral project to keep up appearances in original project quality and ecosystem impact, but gets completely crushed by A in contribution quality, community influence and activity authenticity. A is even more ridiculous: all 14 of its green PRs are thrown at the 4284-star VLMEvalKit as free plug-in labor, its best original project only has 313 st

🔥 Samzong, with 1,614 merged PRs you look like a GitHub beast, but your contribution quality gets crushed by yankay—he’s dropping real core commits into K8s, vllm, Ansible while 57% of your PRs are stuck in your own 50-star Recall echo chamber. Your “top-tier” title is inflated by recycled PRs and a 87-star original that barely passes quality check. Yankay’s 408-star scala-tour alone outshines your entire portfolio—his edge win is built on substance, not spam.

🔥 chenyme edges out sn0wo2 by 11.3 points in this close matchup, but the gap is way more lopsided when you break down the subscores: sn0wo2’s only winning metric, ecosystem impact, is almost entirely made up of trivial doc tweaks for 36k-star repos and slapping his name on org projects, his original project quality is garbage at 9.1/18 with not a single decent original repo to his name, he gets clapped in both contribution quality and community influence, he’s only lucky his ecosystem impact carry

🔥 dai-shi crushes rust with a 24-point gap in a total blowout, and it’s entirely rust’s own fault: he only scored 13.2 out of 18 on original project quality, his highest-starred repo barely hits 10 stars with no decent personal portfolio to speak of, 90% of his PRs are locked into a single jpmobile repo, and he gets outclassed across every metric from contribution quality to ecosystem impact, this gap is 100% deserved.

🔥 chansee97 has been on GitHub for nearly 5 years and only scraped together 39 followers, his most popular repo nova-admin is a Vue admin template with a quality score of just 1, 3 of his 4 high-star external contributions are trivial PRs that only fix typos and tweak styles. He’s so obsessed with using GitHub to package himself as an open source figure for his resume that he’s completely crushed by torvalds in every single dimension. Torvalds, the father of Linux, has a 230k-star linux repo as hi

🔥 Let’s start with the so-called "GOD" tier caicaiis, whose god status is as flimsy as a paper cup: with 284 followers chasing 387 accounts, his open-source bootlicker KPI is calculated to the last decimal. His 12 original repos top out at 34 stars, 73% of his merged PRs are all dumped into the single openinfer-project/openinfer repo to work for free for others’ high-star projects, his original project quality is so barren it’s a desert, and he can’t even pull out a decent flagship project? How da

🔥 Both OSS veterans are GOD-tier, and this dead-even matchup has justjavac scraping a win purely by maxing out the 27-point contribution quality score, while torvalds lags at 22.7 on the same metric — even with his perfect 18-point original project score for Linux, he can’t hide that he treats GitHub like his personal backyard, barely contributes to other repos, and his 310k followers/0 following flex can’t cover that 4.3-point gap. As for justjavac, don’t get cocky: your original project quality

🔥 Both of you are GOD-tier GitHub users with only a 0.9 point gap, this is basically a dead-even matchup, but antfu still scraped a win on activity authenticity. gaogaotiantian, you do have 182 commits to CPython to your name, but your recent merged PRs are full of deleting sponsor logos and tweaking CI configs? C'mon, even the KPIs for CPython core devs include wiping sponsors' messes now? That 0.9 point deduction in activity authenticity is exactly why you lost this round. As for antfu, you're s

🔥 Both of you are GOD-tier OSS veterans, and that 2.5 point gap is literally just the difference between a few drops of squeezed-out fluff—nowhere near a crushing win. luckyyyyy, sure your original project quality is half a step higher than stackia’s, and your miu2d 2D engine is a decent flex, but stackia’s 228 all-substance merged PRs straight up beat you in contribution quality and ecosystem impact. Also, your 8 tokimo-lab org projects combined only have 21 stars? That’s less than a tenth of sta

🔥 The 22.2 point gap basically welds the word 'crush' right onto your face: asperformias, your 10 original projects have a combined total of just 1 star, nearly half of your 96 merged PRs are dumped into the 127-star ghfind, and your long-term contributions to the 140k-star dify and 5k-star dubbo-go add up to only 7. You turned GitHub into an open source outsourcing resume template, and on top of that, you got full-dimension crushed by kohakublueleaf in all six sub-scores, even your community infl