Orbi
Open-source (AGPL-3.0), self-hostable lights-out software factory for the agent era: label a GitHub Issue, get an independently reviewed, merged PR and a tagged release.
DEV TRENDS / DAILY DISCOVERIES
Daily discoveries from developer communities, independent makers, and open source.
Open-source (AGPL-3.0), self-hostable lights-out software factory for the agent era: label a GitHub Issue, get an independently reviewed, merged PR and a tagged release.
AI Car Video 是一款把一张汽车照片变成 AI 汽车短视频的在线工具。最近在 TikTok 和 Reels 上很火的"玩具车变真车"效果:火车窗台上的小汽车穿过隧道,变成在夜路上飞驰的真车。现在不需要模型、不用写提示词、也不用剪辑时间线,上传一张清晰的车辆照片(JPG / PNG / WebP),一键生成约 15 秒的竖屏视频,车身颜色、造型和轮毂在缩放转场中保持一致,下载后就能直接发到抖音、小红书、Reels 或 Shorts。 适合汽车博主、模型玩家、车行和短视频创作者。立即试试:AI Car Video。
OriginalDeco My Tree 是一款免费的线上圣诞许愿树。一分钟创建属于你的圣诞树,把专属链接分享给朋友和家人,每个人都可以挂上一个装饰并留下私密祝福。所有祝福都会保密到 12 月 25 日,然后像树下的礼物一样一起打开。无需下载,手机和电脑都能用,适合异地好友、家人、同学和同事一起过节。立即创建你的圣诞祝福树:Deco My Tree。
OriginalUnofficial Android client for Alma Desktop — pair over Alma's end-to-end encrypted relay
an interactive 3D office for main agents, subagents, and chat delegation, paired with a dashboard for tokens, cache hits, and API-equivalent cost estimates. Go from “who is doing what?” to “where did the tokens go?” in one browser window.
A shared pixel town where your own AI agents explore, fish, read, and meet neighbors. Follow their lives from your browser.
八字和紫微斗数排盘网站,按出生地经度、均时差和各地历史夏令时自动换算真太阳时,免登录排盘;可免费生成 AI 写的命盘报告,前两节免费读,全文 US$9.90,有简繁两版 - 真太阳时换算器
OriginalLocal-first Friday invoice chase ranker — overdue days / 60+ first. MIT. No login, no tracking.
Claude Code / Codex API 中转站,一个账号调用 Claude(Opus 5.5、Sonnet 5.5 等)和 GPT(GPT-6.1、GPT-5.6 等)模型;Claude Code 填 ` 等 OpenAI 兼容工具填 ` 充值 ¥1 = $1,余额不过期,按量计费,无月费 - 中文教程
OriginalGPT代充 / GPT充值|ChatGPT Plus 国内充值指南(2026):ChatGPT Plus 代充怎么选、官方价格、官网订阅 / 应用内购 / 礼品卡 / 代充四种方法和 12 个常见报错,附官方链接与复测记录。作者自营代充商城,文中如实披露,非官方。
OriginalCodebase intelligence for AI and humans: code health scores, auto-generated docs, git analytics, dead code detection, and architectural decisions via MCP.
它把早期个人网站的访问计数器和留言簿组合成一个很克制的交互:序号让人感到“我来过”,一句短留言则把这种痕迹留给下一个人。适合作为极简网页设计、微交互和互联网小实验的体验案例。 当前边界 手机端可看访客序号,留言墙及留言入口目前只在电脑端显示。序号基于浏览器本地存储保存,不应把计数当作严格去重的独立用户统计。 感谢审阅。
Original产品网站:小咖拼豆 我是独立开发者 Owen。把一张图片变成能照着拼的拼豆图纸挺麻烦:要数格子、对色号、算每种颜色要多少颗,打印出来还常常和底板对不上。所以我做了小咖拼豆。 上传 JPG、PNG 或 WebP 图片,就能生成可编辑的拼豆图纸。直接转换免费,在浏览器本地完成,不用登录,比较适合卡通、插画、Logo 和像素图。生成后可以调格数、颜色数量、背景和描边,也能逐格修改。 支持 MARD、Perler(参考色号)、Hama、Artkal 色卡,能看到每种颜色的用量,并导出 CSV 材料清单 填入实测的柱距,就能导出 A4 分板 PDF,附 10 mm 校准标尺;也可以下载 SVG、PNG 另有免费图纸库、图片转像素画和在线像素画工具 真人和宠物照片可以用 AI 模式转换,这个模式需要登录,按积分收费;其他功能都免费。欢迎试用,也欢迎提意见。
Original纳瓦尔先生说“最好的职业就是终生学习者在自由市场中的创造性表达”。 我真的觉得这是一件很酷的事,所以做出了折光社区。 在社区中支持直接部署自己的作品到社区中,可以是游戏也可以是应用,还可以通过短链分享出去。 当然也支持大家直接在网页中生成自己需要的应用。 现在还处于初期,希望能够听到更多的用户反馈,让整个社区生态可以更好的服务到更多的创作者和用户。 网页地址: QQ群: 1092372613
OriginalAutomated vocabulary mining from media.
免费在线工具,用于网申等场景:输入目标大小(如 50 KB),自动把照片压到不超过这个大小的最高画质;也能处理签名(按 px/cm 定尺寸)和 PDF。文件不上传,全部在浏览器本地完成,不用注册。 核心算法开源:
OriginalFree, open-source desktop app for Claude Code and Codex: follow every agent and subagent live, run your own agent organization and automate routines. Also searches your Gemini CLI and Grok history. macOS, Windows, Linux.
精选 AI 资讯信源清单(中英双语)· Curated AI news sources, read daily by Zhiyike
Original常用的窗口挂起来,随点随用。 Your most-used windows, hung up and one click away.
OriginalPaste an arXiv paper and get an interactive course.
An all-in-one email service platform powered by Cloudflare | 基于 Cloudflare 的一站式邮箱服务平台 | Cloudflare Email 邮箱 Mail
Original一款能清垃圾、优化系统、查硬件、管理启动项、查磁盘空间占用、分析蓝屏——等10个模块,外加 34 个系统原生小工具的小巧好用的软件。
OriginalIdeaPanner 是一个给独立开发者用的免费工具(网页 + Chrome 插件),解决两件事:做什么、做完了能不能上线。 1. 找点子:从「已经在赚钱的产品」倒推。TrustMRR 上列了很多用 Stripe 验证过收入的小产品,装上插件之后边刷边标记(跳过 / 先放着 / 深挖),AI 会解释这个产品是做什么的、怎么收钱,并把搜索量、关键词难度、CPC 和排在前面的竞品挂在它下面。动手写代码之前,需求先查过一遍。 2. 上线前体检:贴一个网址,60 秒左右给出 Launch Score。首页跑 39 条规则检查(SEO、隐私条款、移动端 viewport、页面体积等),规则查不了的交给 AI 看,最后只列出最该改的 3 件事,并附上可以直接粘进 Cursor / Claude Code 的修改提示词。 另外有几个不用登录的小工具:SaaS 点子生成器(每个点子都附真实的 Reddit 帖子作来源)、Reddit 各创业板块的月度热帖和抱怨帖整理。 插件和网页版都免费,插件不登录也能用。界面目前是英文。 Chrome 插件: 网页版: 作者是一个白天上班、下班做独立产品的前端开发。
OriginalPasteFerry (隔空贴): clipboard and file sync across macOS, Windows and Linux. Direct-first, end-to-end encrypted. Optional official access: no server to host. Independent self-hosting remains available. 开启官方服务即可跨网络同步。
Original巡影 · 开源 AI 影视自动追更与多网盘媒体库整理工具 / Open-source AI agent for cloud drive media tracking & NAS (夸克 Quark / UC / 115 / 123 / 光鸭 Guangya / 天翼 Tianyi)
Original红果短剧的非官方匿名 SDK,提供 Go、Node.js、Android、iOS、React Native 和 Flutter 版本,附命令行演示程序 ohg CLI
OriginalA native, lightweight Git client for macOS. Manage repositories and worktrees, review diffs, and commit changes in one window.
面向医疗 AI 智能体的失效审计基准:在坏数据、坏工具、坏时序下,度量临床大模型智能体的失败方式、错误传播与不确定性表达。主实验用 Synthea 纯合成患者,零合规风险、可公开复现。
OriginalAn anime aggregation, parsing, and playback app
One local-first desktop workspace for every AI coding agent — Codex, Claude, OpenCode and your own models.
追忆似水年华是127万词 (中文约200万字) 的文学巨著, 好几份世界文学top榜单都是仅次于唐吉坷德的文学巅峰. 然而其在简体中文只有一份完整译本, 是40年前15人合译的. 而本项目 在保留原著文学性的同时 比15人译本更加流畅, 没有错译, 注解丰富 这部文学巨著的翻译工作意味着AI在文本翻译上已经远远甩开人类一大截 (工时对比: 15人三年 vs 人类 + AI 5天 vs 单人译者 规划10年); 并且 因为比先前的译本更加流畅, 这部文学经典也会走进更多人的视野, 原本"普鲁斯特难读 读不下去"的印象也会减弱 本项目使用ai翻译(gpt-6.1 sol + qwen3.8-max + deepseek-v4.1-flash), 累计消耗100亿+token, 耗时约5天, 进行多轮审校, 耗资近100元巨款 (有些token是白送的) (往多了算也不超过800元) CC BY-NC-SA 4.0 开源 + 开放PR / Issue 工作流: 原文分段 逐段翻译 -> 逐段校对 (到后面光术语表都占20M Token了) 分章节 全章校对1 分章节 全章校对2+加注释 分章节 全章再次排查查错译 分章节 补充更多双关语的注释
OriginalCloudflare Pages 多站点管理后台:评论、浏览量、会员、R2 文件 AI 网关。One serverless dashboard for all your Cloudflare sites.
Original阮老师好,自荐一个面向开发者的开源项目发现网站:HotRepo。 看 GitHub Trending 时,常常需要逐个打开仓库、阅读 README,才能判断一个项目是否值得继续了解。HotRepo 把热门项目榜单和中英双语 AI 解读放在一起,希望减少初步筛选项目的时间。 目前可以: 浏览 GitHub 热门项目和新项目榜,查看项目简介、语言等信息。 阅读项目的中英双语 AI 解读,再跳转到原始仓库深入了解。 按分类发现项目,回看历史日榜和月报。 解读由 AI 生成,适合用作了解项目的起点;具体功能、使用限制和许可证仍应以原仓库为准。
OriginalThe fast tier of the wise-men council as its own skill: three Claude subagents down three reasoning paths, one checked decision memo in a third of the full council's time. Ships its own blind eval.
原生 macOS 下载工具(SwiftUI)。统一管理磁力 / 种子 / ed2k、HTTP 与网盘直链,以及 YouTube、哔哩哔哩、抖音、小红书等平台的视频下载;支持浏览器扩展接管、边下边播、断点续传与 PikPak 云盘离线。基于 Gopeed 与 yt-dlp,GPL-3.0 开源。
Original面向影视创作者与开发者的 Mac 清理工具。识别 60 余款剪辑、设计与 AI 工具的缓存,一键清理,可撤销,不影响项目、素材与对话。
OriginalNative Mac dictation in a 12 MB download. Free Local mode with Whisper + Parakeet, reproducible benchmarks, optional Cloud. MIT-licensed client.
Documentation and setup guide for the MCP interface built into the UniPublisher desktop application (1.6.8+). Not a standalone server.
Stanford CS329Z Engineering AI Agents 中文教程:19 讲与 4 项实践,涵盖 LLM、RAG、工具调用、Agent 记忆、多智能体、评测与安全。完整 Markdown 和在线/离线 HTML。Chinese course notes and hands-on guide.
OriginalOpen-source StayFocusd alternative and website blocker for Chrome. Set daily time limits, block distractions, and track screen time locally. No accounts or telemetry.
The design vocabulary for your coding agent. Steer with polish and bolder, review with critique and audit, catch AI slop before it ships.
Vibe code with pi in a lightweight, real IDE that customises itself around the way you work — The Vibe You Need
Open-source Grok Bot alternative with a virtual machine that bots can use
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
Open-source AI coworkers that each get a computer of their own: a browser, files and tools, with every action decided before it happens and recorded after. Bring any AG-UI agent.
Turn Claude Code into a full game dev studio — 49 AI agents, 72 workflow skills, and a complete coordination system mirroring real studio hierarchy.
octocat: 分享 GitHub 上有趣、入门级的开源项目。Share interesting, entry-level open source projects on GitHub.
OriginalOpen-source personal AI assistant & Agent Harness. Plans tasks, runs tools and skills, self-evolves with memory and knowledge. Multi-agent, multi-model, multi-channel. Lightweight, extensible, one-line install.
The voice-first assistant for small business owners
The music layerfor your AI stack
GitHub and Linear issues, right inside your Codex
The AI labs want to own you. Make them answer to you.
Easiest way to build hardware gadgets using plain English
Better data. Smarter robots. Repeat.
Email that groups your inbox by contact
JST-2: 3D scenes, characters and motion from one model
JIMMY is an AI agent that runs on your PC
An intelligent email inbox that works for you
Cloud browser for AI agents and tough sites
Automate any folder on your Mac with if-this-then-that rules
New AI agent, 1,650+ apps, Skills and smart model routing
Build and publish real apps without leaving ChatGPT
text-to-speech built for voice agents
Why is it useful? - Present slides without sharing the screen and have every one be on the same page. - Present together with other cohosts and all control the current page. - Other features such as presentation collab features. Why should I not just vibe code my own solution? - Its a hassle to maintain a version online.
Let's take life a little slower shall we?
US federal income tax forms and instructions as cited, executable, machine-checkable data.
God's chosen Coding Harness
Hey everyone! I couldn't find an app to organize trips that gave me all the info I wanted to schedule each stop per day, with travel times, leeway, etc, so I made Wherewhen. The basic differentiator is that you can set various constraints (e.g. "I want to spend 2 hours in the park", or "I have a dinner reservation at 6"), and Wherewhen will automatically adjust the rest of the times to tell you when to go/leave each place, including travel. I've tried to make it mobile-friendly but travel planning is pretty complicated and it's a bit clunky on mobile. Feedback is appreciated!
Email inboxes for AI agents, in the EU, and never trained on your mail’s content. Threads, webhooks, IMAP, MCP and the Firewall, with your own domain free.
I burned about $500 for this fun project: Small Python. Initial PoC used 3 resets of Codex in one session but all later development happens in Claude Code, also in one Opus 5.5 session over three days. The outcome is a quickjs-like implementation for Python. Almost full Python 3.15 language features plus builtins are covered. No Lib/ support. Great for embedding scenarios. No rust. Just plain C11 code. is here: https://github.com/soasme/small-python/ Ask me anything.
Jotcal is a family calendar you glance at, not one you scroll through. On iOS and Android, with handwriting recognition, live weather, and adaptive layouts.
Hello HN, I've been working on a new Web engine, build around two ideas: - use formal methods from day one. - make it modular. The modularity is orientated towards the following goals: - enabling shipping anywhere, including on IOS in the EU as per Apple’s guidelines. - only write code for stuff the OS or application platform cannot do for you: keep binary size to a minimum and achieve native performance. - developer choice: bring what you want, and only build what you need. For example you can choose a JS engine like v8 or Boa, or just bring your own. The entire graphics pipeline consists of components whose implementation can be swapped-out, with currently a relatively fast path on Mac integrating with Core Animation and a slower cross platform path. Project is at https://github.com/gterzian/formal-web Intro essay at: https://medium.com/@polyglot_factotum/how-i-am-building-a-we... Please let me know what you think.
Hello all, Ive been building Fabivo, a free furniture planner that runs in the browser. You sketch it simply by drawing it, (by far my favorite feature of this app), clicking for boards or just selecting some starting designs. But what i think separates fabivo from others its the image to model ai assistant, which took me a looong time to develop and get some quirks fixed, thanks god latest models are focusing on 3d, thats been helping too! The photo part is an image to JSON CAD harness: the model returns a parametric furniture model in mm that the app can build and edit. That harness is also what i benchmark, and honestly its the reason the app exists. I wanted to see if a photo could turn into a plan i could take to the saw. Also finetuned some image models and found interesting stuff, that was a first. My intention was to basically reuse the drawing feature to convert a 2d line sketch from image models like gpt image / nano banana to my parametric model, which "worked", and was actually thinking of using that for prod, but then cheaper/better models at 3d launched, but anyways, i learned a lot! Im getting better performance from my harness than from plain 3js, tho this is anecdotally and havent actually benchmarked it... Happy to answer anything, especially about the harness or the cut order. edit: Adding the numbers, missed them. 17 photos of real furniture, three models, same harness, medium reasoning, same tools. Scored on how close the returned structure and boards are to the gold model, so its my own set. (tho i have a hard time making a good scoring metric... just usually go by look!) Model (medium reasoning) Avg. time / photo Avg. input tokens / photo Avg. output tokens / photo Avg. tool calls / photo GPT-6.1 Terra 9.5 s 30,409 632 1.06 Claude Sonnet 5.5 13.6 s 50,167 1,592 1.18 Claude Haiku 5.5 34.9 s 66,786 6,658 2.35 Terra was both cheaper per photo and the most consistent to my likings. Sonnet was close on quality, Haiku needed way more tool calls to get there.
A marketplace for CS2 trading, built as an university course project. Built to trade. Made to stand out.
Hey HN, I’m Abhishek. I'm building Opslane, a tool that finds reliability issues in your code. On every PR, it spins up your entire app and then tries to find weak spots that would break in production. https://youtu.be/47phxkdh6f0 Example: We tested Opslane on Lago (open-source billing) and found a bug where customers get double charged. Opslane made Stripe take the money and then return a 500. Lago marked the payment failed. A retry from the dashboard charged the customer again. https://github.com/getlago/lago-api/issues/6903 Why I built this I used to work on reliability at Robinhood. We used to run chaos testing in our staging environments: inject faults in various parts of the system and see what breaks. We found issues like Kafka or Redis being unavailable for a few seconds could take the whole app down. But that kind of testing was expensive. It needed a well maintained staging environment. Engineers needed to know what to test. Most teams never actually did it. Over the last year, coding agents have gone writing some of my code to almost all my code now. Teams are shipping way more code than they were 6 months ago. Code review tools read the diff, but a lot of reliability bugs aren't in the diff. They're in how your code behaves when Stripe, Redis or the network misbehaves. And e2e tests usually mock those services out and test the happy path. Agents have made this type of testing a lot cheaper. And we will see more of this (kinda like security scanning / testing) How it works A full copy of your app per PR: Postgres, Redis, Kafka, SQS, S3. Working copies of your third-party dependencies: Stripe, Slack, Gmail and more. Traffic goes through our proxy, which lets us inject failures. The agent picks what to test: It reads every PR and writes down invariants that need to hold (for e.g. a customer should never be charged twice). It then injects faults likely to break the invariants. Bugs we've found We ran Opslane on some popular open-source apps and here are some of the bugs we have found: 1/ Twenty: deleting a workspace while Stripe is failing locks everyone out, and billing keeps going. Twenty merged a fix. ( https://github.com/twentyhq/twenty/issues/28043 ) 2/ Dub: a restart in the middle of a payout paid a partner $240 for $120 owed. ( https://github.com/dubinc/dub.link/issues/4707 ) Full list: opslane.com/open-source Opslane is pretty early - we have a couple of early customers. We have found bugs in their system that their other code review tools missed. Would love to get feedback from folks and how they are seeing the reliability of their systems evolve over time with agents.
A fun, personal project - AI did a pretty good job.
Hi HN, I’m Romain, a designer from France. I’m trying to build a tool to make art with code. This is the first piece. Drop any image and the particles take its colours. You can record a clip of the result. Built with three.js. Free and open source Feedback welcome.
LFM3 video chat
我的美股散落在好几个券商里, 一直找不到一个地方能看清楚: 总共持有什么, 今年到底赚了多少, 跑没跑赢大盘啊。券商 App 只看得到自己那一家, 记账软件要手动录每一笔交易。所以干脆自己写了一个。 - 支持各种券商 api 连接,不用手动输入(当然也支持自己手动导入 csv )。 - 买入点卖出点都非常清晰显示在线上,还有金额细节,成本价都算的清清楚楚的。 - 账户数据都是自动算,合并算分开算,非常直观。 - 真实收益: 同时算时间加权 (TWR) 和资金加权 (MWR) 收益,可以和 SPY 、QQQ 等基准对比,不会被定投入金带偏。 - 收益来源: 按标的拆分收益,看清今年的钱是哪几只赚的。 - ETF 穿透: 把所有账户里的 ETF 拆开,算出合并后对单只股票的真实敞口。比如 QQQ 、VGT 、VOO 叠在一起,你实际持有多少 NVDA 。 - 分红:已到账的分红和今年的预期分红。 - 研究工具: 成交量分布、期权持仓墙 (OI walls)、市场情绪仪表盘,还能看佩洛西、段永平、巴菲特等知名投资者公开披露的持仓。 - AI: 内置 AI 助手,用你自己的 AI;还提供 MCP server,可以在 Claude Code 或 Codex 里直接问问题。 - 支持 A 股!可以搜索添加 几个设计原则: - 只读,不能下单。 - 计算都在手机本地完成。 - 不包含任何投资建议哈。 **App Store**: **官网**: 我已经修了好久,bug 已经不多了,但是一个人 vibe 开发,肯定有不少问题,欢迎拍砖。特别想知道大家还在用哪些券商是连不上的。
Original各位 V 友大家好,我是独立开发者。 家里买过 **兄弟( Brother DCP-B7520DW / B7535DW / T420W / T720DW / L2535DW /L1848W 等)**,或者惠普、佳能、爱普生、奔图无线打印机的朋友,大概率都被以下三个问题折磨过: 1. **手机打出来的字远不如电脑清楚(严重锯齿、笔画发虚)**:明明同一台激光/喷墨打印机,用电脑插 USB 线打出来字迹如刀刻般锐利;但一换到手机官方 App (比如 `Brother iPrint&Scan`)、微信小程序或者安卓系统默认打印服务,打出来的汉字边缘全是阶梯状马赛克锯齿,细横画断裂发虚,小字甚至糊成一团。 2. **手机直接打 Word (`.docx`) 必定变字体、乱排版**:电脑上排得好好的宋体、楷体、仿宋、小标宋公文或数学试卷,传到手机微信里直接打开打印,全部被迫坍塌成同一种无衬线黑体,粗体黏连、行距错位、莫名其妙多出一页空白。 3. **异地远程打印配置太劝退**:人在公司上班或出差,老师在群里发了卷子要打,家里老人不会弄。如果在软路由或 NAS 上折腾 CUPS + 找 Linux 驱动 + 配 frp/Tailscale 内网穿透,不仅步骤繁琐,家里人的手机还得跟着挂穿透客户端,体验非常割裂。 一怒之下,我耗时几个月手搓了这套 **「鹰眼打印 (HawkEye Print)」** —— 把**原生 600/1200 DPI 1-Bit 激光点阵渲染引擎**、**30 款 Windows 原版字库**和**免公网 IP 的多平台远程打印网关**装进了手机里。 - 🌐 **官方网站 & 全平台网关直下**: --- ### 一、 底层技术拆解:我是怎么根治“锯齿发虚”和“Word 变字体”的? #### 1. 绕过系统 150~300 DPI 灰度缩水,原生 600/1200 DPI 1-Bit 点阵流直驱 为什么普通手机打印会发虚有锯齿? 因为安卓默认打印框架与大多数小程序在发送任务时,会先将页面光栅化缩水成 **150~300 DPI 的 RGB/灰度位图**,打印机收到低清灰度图后再做二次**半色调网点抖动( Halftone Dithering )**——原本应该 100% 纯黑落粉的笔画边缘,全变成了散落的灰点与马赛克阶梯。 **鹰眼打印的做法**: - 彻底绕过系统低清打印转换管线,直接在手机本地以 **原生 600 DPI / 1200 DPI** 进行超高清矢量光栅化; - 通过自适应二值化阈值算法,直接生成 **1-Bit 纯血激光像素点阵流( PJL / PCL / Raster )**,通过局域网 `RAW (9100)` 或 `IPP (631)` 端口直驱打印机; - 打印机不再做任何二次灰阶抖动,落纸效果与电脑插 USB 数据线直连 **100% 完全一致**。App 内置了「彩色原稿 / 黑白二值化实效」双模切换,打印前可以直接在手机上无极放大 10 倍检查每一根笔画的像素边缘。 #### 2. 自研 Word 矢量排版引擎 + 30 款 Windows 原版字体按需秒配( 100% 本地解析零上传) 普通安卓手机系统只内置了单一的无衬线黑体(`DroidSansFallback` / `NotoSansCJK`),缺少电脑 Windows 上的常用字库。 **鹰眼打印的做法**: - 内置自研的 `.docx` 矢量排版解析引擎,文档 **100% 在手机本地解析渲染(零上传云端服务器,绝不泄露合同、公文与试卷隐私)**; - 当解析到文档使用了手机缺失的字体(如 **宋体、楷体、仿宋_GB2312 、方正小标宋、黑体、微软雅黑、等线、华文行楷、Times New Roman 、Cambria Math 数学公式** 等 30 款 Windows 原版字重)时,仅从云端毫秒级按需拉取对应的轻量字体包并永久缓存到本地; - 真正实现页边距、字间距、行距、中西文混排与真粗体的 **1:1 电脑级还原**。 --- ### 二、100% 永久免费的四大「家中远程打印网关」(无需公网 IP · 免配内网穿透) 为了解决在外地用 4G/5G 远程控制家里打印机的问题,我把**「家中打印网关」做成了 100% 永久免费、不限时长、不限接收页数**,无需公网 IP ,也不用装任何 VPN ,只需记下网关生成的 **6 位数字配对码**,外地手机输码即可秒连出纸(首次连上后自动记忆,下次点一下 3 秒出纸): 1. **🛜 OpenWrt / iStoreOS 软路由插件( 365 天×24 小时零功耗值守)** - 提供了全架构通用的 `luci-app-hawkeye-gateway_1.2.6-1_all.ipk`( x86 / ARM 均可用,体积小巧,彻底告别臃肿的 CUPS ); - 上传 ipk 安装后,在 LuCI ` [服务] -> [鹰眼打印] ` 勾选启用,填入打印机局域网 IP (如 `192.168.1.150`),即可看到 6 位配对码。 2. **🗄️ 群晖 Synology / 飞牛 fnOS / 绿联等 NAS 部署( 24 小时无人值守)** - **群晖免 Docker 10 秒部署法(最稳最快)**:无需折腾 Docker 镜像源!解压 `hawkeye-gateway-nas.zip` 到 NAS 后,直接在群晖 ` [控制面板] -> [任务计划] ` 新增一条 `root` 自定义脚本运行即可: ```bash nohup python3 /volume1/docker/hawkeye/lasersharp_gateway.py --headless --web-port 19180 >/dev/null 2>&1 & ``` - **飞牛 fnOS / Docker Compose 部署法**:打开飞牛 ` [ Docker ] -> [ Compose ] -> [新增项目] ` 选中解压目录即可一键启动; - 部署后浏览器访问 ` NAS_IP:19180` 即可打开可视化 Web 控制台查看 6 位配对码并绑定打印机 IP 。 3. **💻 Windows / Mac 电脑一键网关** - 下载解压 `hawkeye-gateway-pc.zip`,Windows 提供 `Windows 后台静默运行(无黑框).vbs`(可丢进 `shell:startup` 开机无感自启),Mac 双击 `.command` 即可运行。 4. **📱 家里闲置旧安卓机 / 长辈手机直接当网关(零硬件门槛)** - 家里淘汰的旧安卓手机( Android 7.0~15 均可)装上鹰眼打印 App ,切到 `「🏠 作为家中网关」` 页面放着,就能免费充当打印网关。 三、 针对打试卷与日常办公的几个顺手功能 🧹 拍照试卷一键净白去底灰 & 剥离斜向水印:手机拍下来的试卷常有阴影黑边或机构彩色斜水印,开启「去水印 / 防粘连 / 加深」与「红黑双色分离」档位,自动剥离底灰,省墨不卡纸。 📘 单面打印机「手动双面打印傻瓜向导」:家里的单面机想打双面?软件会自动先打完全部奇数页 -> 弹出直观的翻纸方向图解 -> 点击继续后自动按 [逆序] 打印偶数页,打完直接装订,再也不用动脑猜正反面和页码顺序。 🖨️ 广泛兼容主流机型: 兄弟( Brother )全系:DCP-T420W/T425W/T426W 、DCP-T720DW/T725DW 、DCP-T820DW 、MFC-T920DW 、DCP-B7520DW/B7535DW 、HL-B2080DW 、DCP-L2535DW/L2550DW 、HL-L2350DW 、MFC-L2710DW 等; 其他主流品牌:支持 IPP (631) 或 RAW (9100) 协议的惠普( HP )、佳能( Canon )、爱普生( Epson )、奔图( Pantum )、联想、施乐、京瓷等 Wi-Fi / 局域网打印机。 四、 关于免费额度与 V 友专属福利 🎁 永久免费部分:所有平台的「家中远程打印网关」 100% 永久免费;每台手机每天赠送 局域网免费 3 页 + 新机赠送 远程打印免费 10 页(足够大家完整测试打印机连接、对比 1200DPI 刀刻画质与异地出纸效果)。 Pro 会员定价:觉得确实解决了痛点再考虑支持,年付仅 ¥12/年,永久买断 ¥29.9 ~ ¥59.9 (一机一码硬件级绑定,卸载重装联网 1 秒自动恢复会员)。 🎉 V2EX 专属送码福利: 非常感谢大家帮我测试各种品牌的打印机与 NAS/软路由环境! 回帖留下你的 [打印机型号 + 准备用什么设备当网关(如飞牛/群晖/OpenWrt/旧手机)] ,并附上 App 会员中心显示的 [ 10 位本机设备码(如 8F3A9-C12B4 )] ,前 100 层楼我直接后台为你免费激活 Pro 会员! 🌐 官网地址(含 APK / NAS 包 / OpenWrt ipk / PC 网关直下):
OriginalA self-hosted 18+ personal media library for organizing and playing your own videos.
Open-source macOS Pinyin input method with Chinese-to-English translation using Sign in with ChatGPT.
打开就能玩,不用下载也不用注册,电脑和手机(横屏)都行。 ### 怎么玩 - 每局随机分成两队:🎃 南瓜头(最多 26 人)和 👻 幽灵(最多 4 只) - 南瓜头要从西边的 START 跑到东边的 FINISH ,跑到就算逃脱 - 幽灵要**挥爪**才能抓到人,光碰到不算。爪子有攻击范围和冷却,也打不穿墙 - 两队都能冲刺( Space / 右键),场上的糖果吃了能加速 - 一半南瓜头逃脱就是南瓜头赢,否则幽灵赢 ### 技术上的一些东西 - 前端用 three.js ,地图是用 Blender 做的 - 南瓜头角色、幽灵:都是在 Blender 里用 Python 脚本程序化生成的,包括动画 - 糖果、金币:也是脚本生成的( - 吸血鬼翅膀(转盘奖品):用的是 BlenderKit 上 Joachim Bornemann 的 "Count Dracula's Wings",Royalty Free 授权,减面到约 2.4k 三角面,再把贴图烘焙成顶点色。 音乐和音效 - 背景音乐:两首都来自 Pixabay 。一首是 sigmamusicart 的 "Halloween Background Music"(#597995 ),另一首是 nastelbom 的 "Halloween Horror"(#616999 )。 - 大部分音效:来自 Kenney 的 CC0 素材包,包括糖果、倒计时、GO 、到达区域、被抓、抓人、逃脱、胜负、聊天提示音。 - 后端跑在 Cloudflare Workers + Durable Objects 上:一个 LobbyDO 负责匹配,每个房间一个 RoomDO ,跑 20Hz 的游戏循环 一局就几分钟, 适合摸鱼, 大家可以试试给点玩法建议 ↓↓↓↓↓↓游戏地址↓↓↓↓↓↓ 🎃 Join me in Trick or Treat Run! Room 1031 看后台最近有个网站发现了这个游戏 带了一些国内的玩家进来
OriginalmacOS-style Stage Manager for Windows 10/11: keep the current app on stage and group the rest into a sidebar, one per monitor. Pure Go, no cgo, no WebView.
大家好,分享一下我做的 HotRepo: 它是一个 GitHub 项目发现站,把热门项目榜单和中英双语 AI 解读放在一起,想让大家在逐个打开仓库、细读 README 之前,先了解项目大概做什么,再决定是否深入研究。 目前主要有这些功能: - GitHub 热门项目和新项目榜,展示项目简介、语言等信息。 - 中英双语 AI 项目解读,并保留原仓库入口。 - 分类浏览、历史日榜和月报,方便回看之前出现过的项目。 技术上用了 Nuxt SSR + FastAPI ,同时提供中文和英文页面。 AI 解读也有局限:可能遗漏信息或理解有误,不能替代项目文档和实际测试。具体功能、许可证、部署条件等,还是要以原仓库为准。 欢迎大家试用,也想听听大家筛选开源项目时最在意哪些信息,以及现在的榜单和解读还有什么不方便的地方。
Original大家好,我是 [BibGenie]( ) 的开发者,来分享一下自己做了一年多的东西,也想听听 V 站里还在用 Zotero 看文献的朋友的真实反馈。 ## 背景说明 用 Zotero 的应该都知道,它是科研圈最主流的开源文献管理工具,但本身没有任何 AI 能力。我自己读论文时有个很具体的痛点:把论文丢给 AI 问,答案说得头头是道,但我根本不知道它是从哪篇、哪一页得出的结论,最后还得自己回去翻 PDF 核对,费时费力。 所以我做了 BibGenie (去年叫 Zotero Copilot ,但因为 Zotero 商标的问题改了名),一个原生跑在 Zotero 里面的 AI 插件,侧边栏形态,可以直接访问官网 查看首页的展示视频,视频会更加直观。 迭代了一年多,目前更新到 v0.9.4 版本了,核心功能都已经开发完成了,今天来 V2EX 求一点反馈 (顺便吐槽一下,现在很多独立开发者只顾哐哐开发新产品,却不管维护,我也是深受其害。) ## 核心功能大概是这几块: 1. 直接和自己的 Zotero 整个文库对话。 不用导出文件、不用把 PDF 一个个上传到网页工具。在 Zotero 里面直接用 @ 或者右键 Add to Chat ,就能把条目、PDF 的某几页、EPUB 章节、批注、笔记、标签、整个集合拉进同一个对话里,多篇论文一起对比之类的。 2. **引用可追溯到文本块。** 这是我最在意、也花时间最多的功能。AI 回答里给出的引用,点一下能跳回 PDF 对应页,并高亮出它引用的那个文本块; EPUB 、网页快照也是一样定位到段落。科研场景下,我觉得这个功能真是刚需。 3. **本地优先,支持 BYOK**。聊天记录存在本地 Zotero 数据库里,不上传服务器;文库的语义索引也是在本地建的。模型这边三条路随便选:自带 Key ( OpenAI 、Claude 、DeepSeek 、OpenRouter 、各种兼容接口都支持,请求直连服务商,不消耗额度);本地模型( Ollama 、LM Studio ,聊天和 embedding 都配成本地的就可以完全不出本机);或者用 bibgenie 官方模型开箱即用,什么都不用配置。 4. **内置了一个 Zotero MCP Server**。这个估计 V 站的朋友会比较感兴趣。插件里直接起了个本地 MCP Server ( Streamable HTTP ,长这样:127.0.0.1:23119/bibgenie/mcp ),不用再单独装什么 Python 服务。Claude Code 、Claude Desktop 、Cursor 、Cherry Studio 连上之后,可以直接检索我的 Zotero 文库、做语义搜索、读 PDF 附件(页面还能渲染成图片,看图表和扫描页没问题)、读笔记和快照,也能在客户端自己的权限确认下建笔记、整理集合。 5. 帮着整理 Zotero 文库。 可以让它整理集合、清洗标签、核查元数据(时间长了,zotero 文献库的标签,分类之类的都乱了,可以直接来整理)。元数据是交叉比对 arXiv 、Semantic Scholar 、OpenAlex 、DBLP 、OpenReview 这几个来源的。所有改动先出一张 diff 预览卡,点批准了才写进去。 ## 关于定价 所有的功能都是免费的,需要购买的只有模型的 token 。免费用户每月送 200 credits ,Pro 是 $9.99/月,有 10000 credits ( Pro 肯定够用了!!)。现在还有个 $199 的 Lifetime Pro ,v1.0 之后会下架(当然已经买了的不受任何影响,下架只针对新购),独立开发的小产品确实长期扛不住终身制, 前期已经倒贴了不少了。 官方模型目前是通过 OpenRouter 接入的,我在成本价上只加 7% 用来覆盖服务器和其他支出,目前没打算靠 token 差价赚钱(毕竟也支持自定义模型)。 官网和文档在这: 文档: 最后想认真求一波反馈,尤其是平时用 Zotero 的朋友。批评也完全欢迎,这东西的不少功能就是被用户骂出来的。感谢阅读。 (很抱歉,没有福利带给大家,毕竟这个产品是免费的,相信 V 友都有能力接入自定义模来使用,如果确实有想订阅的,可以留下邮箱,我给你 8 折优惠券。 当然,有 llm 服务商愿意支援 token 网页是欢迎的。)
OriginalTry another keyword or category.
21 repositories were already featured in the last 3 reports, so they are not repeated in today’s discoveries.
DeepTutor: Lifelong Personalized Tutoring. .
LIST OF ALL MY JAILBREAKS
No project description yet.
DEV TRENDS / COMMUNITY
Discuss this edition
Sign in with GitHub to share a thought, question, or useful follow-up.