
WiseReply – Write better replies faster
Instantly transform rough thoughts into professional replies.
@anandkumar_dev · X
The full gallery
Tech stack
33 projects

Instantly transform rough thoughts into professional replies.
@anandkumar_dev · X

Create, send, and sign client proposals with AI in minutes.
sahil423 · HN
18y/O buidling an AI proposal OS and generator

Set of design tools that integrate with your other software.
@aviralgarg05 · X
Here's what you get on One toolkit. Every design task.

Publish plans from your terminal using Claude Code workflows and AI.
@discorev · X
I've been working on - Using gpt-5.6-sol and Fable in Claude Code via CLIProxyAPI. I've found using these two together with Claude Code Workflows is a killer combination

AI agent that generates comparable product concepts from ideas, sketches, images, and constraints.
@hedge8og · X
4月23日,我开始用 Codex 搭自己的产品。到现在,满3个月了。 截至7月19日,Codex 页面显示247.1亿个 Token,连续使用54天。标题里的“250亿”是取整,准确数字以截图为准。 这三个月,我把 做上线了,产品介绍视频改了很多版,也借助 Codex 做了不少以前一个人很难完成的工作。 具体一点:我用它整理过 BP、导出过 PDF、做过 PPT 工作稿,也把调研、文档、图片、浏览器操作、网站、视频和邮件这些常见 Skills 跑过一圈。有些后来真的进入了工作,有些试完就放下了。 现在回头看,最大的收获不是“我做过这么多种东西”,而是我越来越能分清:哪些能力能进入真实工作,哪些只是演示起来很厉害。 这些数字确实挺夸张。说实话,我自己看到也会觉得:是不是做了很多事? 但如果现在让我总结,我不会把它们当成绩。它们只能证明我很投入,蹬得快冒烟了,不能证明产品已经有价值,更不能证明商业已经成立。 我大学辍学,最高学历是高中,之前一直在制造业工作,也不会写代码。写这些不是为了卖惨,更不是想证明学历没用,只是先把我的真实起点说清楚。 2月 OpenClaw 火起来以后,我开始认真试各种 AI 工具,想先解决工作里那些重复、又不得不做的事情。 这期间还有一次,客户马上要我提案,但公司的设计师手上也有急活。我用 ChatGPT 做了产品效果图,最后拿到了订单。 老板后来跟我说:“早知道你可以帮我做这些,那些外包设计早就可以不用买了。” 到了4月,我又用扣子搭了发货模板和记账系统。原来需要手工对比核验的条形码、外箱码,我也开始想办法交给工具处理。 这张订单图只能证明订单记录是真实存在的,不能单靠它证明订单一定是 AI 带来的,更不是 Ohwait 的客户案例。AI 参与了产品效果图制作,之后拿到订单,是我对那次真实工作过程的记录。 那一刻让我兴奋的,其实不是发现了什么创业机会,而是我第一次觉得,模型真的可以成为人的“数字外骨骼”。 回头看这三个月,我现在有5个比较确定的结论。 ### 1. AI最好先从一个真问题开始 发货、记账、核验、临时要效果图,这些都不是什么宏大命题,但它们有一个共同点:问题原本就存在,而且有人正在为它花时间、花钱或者承担压力。 我现在判断一个 AI 工具有没有价值,还是看两件事:它有没有少让我重复

Upload photos and describe edits to remove objects, change backgrounds, or combine image references.
Nano Banana AI — AI 图像生成和编辑工具(免费在线),2-5 秒生成高质量图像,支持文生图和图像编辑

Score business ideas on market viability, competition, and demand.
@buildwithmaya · X
Hi Rashka I am building

Create landing pages to test startup ideas and measure email conversions.
@joffreyio · X

Record a 60-second pitch and get AI feedback on hook, clarity, and confidence with rewrites.
@Ethan_Cohen07 · X
, feel free to reach out to discuss ideas

Tailor your LaTeX resume to job postings through guided AI conversation with version history.
Sriram1729 · HN
Tampa, FL | Remote (US Eastern) | Contract, ~20 hrs/week (evenings + weekends, around a full-time job)I'm a full-stack + AI engineer, 5 years in. Looking for one side engagement I can actually commit to — nights and weekends, not a flaky "I'll get to it" thing. Short version of what I'm good for: give me a problem and I'll hand back something deployed. I do the whole thing — backend, frontend, the data model, the infra, the deploy. You don't need to spec it out for me or check my work line by li

Refine ideas through logical reasoning and publish them to a community of verified knowledge.
samcymbaluk · HN
Axiom toolkit – Make your ideas precise enough to be proven wrong

Manage academic conferences: submit and review papers, handle scheduling and registration.
@harryjwang · X
I've used many of the major conference systems. Plenty still have docs as an unsearchable static site — or worse, as PDFs that stopped tracking the product three versions ago. Somehow, still true in 2026. We took an AI-first path with instead. Our docs are generated from the codebase itself. A custom Claude Code skill writes each page and retakes the screenshots whenever a feature changes — so the docs don't drift from the product. Every page still gets reviewed by a human before it ships. Now you can talk to them, too. We just shipped "Ask AI" over our docs: This feature also turned out to be a nice teaching case for agentic RAG vs. classic RAG: • Classic RAG: chunk → embed → vector DB → retrieve once, by fixed rules • Agentic RAG: hand the model a search tool and let it drive — write a query, read the results, rewrite, retry At more than 100 pages, still no vector database needed. And every answer cites the exact page it came from. #Age