
Poincaré · adaptive liquidity
An adaptive AMM that detects real price trends and adjusts curves to protect liquidity providers.
@Prakhar_3010 · X
The full gallery
Tech stack
60 projects

An adaptive AMM that detects real price trends and adjusts curves to protect liquidity providers.
@Prakhar_3010 · X

Live dashboard showing an AI-agent company's ARR, MRR, tasks, messages, and activity in real-time.
pro_methe5 · HN
I scrape an $8.5M-ARR company run by AI agents and chart it live

Accelerate stock research and automated asset tracking. Parse SEC filings in seconds, track multi-currency portfolios in real time, and access 70,000+ global stocks, ETFs, crypto &
@stockyexplorer · X

US stock trading signal platform analyzing market data to identify trading opportunities and signals.
@MEJ50749 · X
用 @dappOS_com coding 做了一个美股信号网站: 它和普通美股行情网站最大的区别是:普通网站更多是“给你看行情”,而这个网站更像一个“辅助交易工作台”。 大多数美股网站会把价格、涨跌幅、新闻、财报、K 线分散展示出来,信息很多,但用户还要自己判断: 哪只股票值得看? 现在是什么信号? 这个信号强不强? 有没有触发提醒? 后续要不要复盘? 我这个站的核心思路不是堆信息,而是把交易前的观察流程做成工具: 1. 市场总览不是单纯看涨跌 首页会把美股标的放进观察池里,比如 AAPL、NVDA、MSFT、GOOGL、AMZN、META、TSLA 等,不只是显示价格,还会结合涨跌幅、成交量、市值、PE/PB、行业、52 周区间等信息,让我先快速判断哪些票值得继续看。 2. 最重要的是信号扫描器 这个功能是普通行情站很少直接做成工作流的地方。 网站里有“智能信号扫描器”,可以扫描技术形态,并把信号归类成突破、反转、动量、超卖、超买、金叉、死叉、背离等类型。 比如: NVDA 出现强势突破、量能放大、均线多头排列; AAPL 出现 MACD 金叉、RSI 超卖反弹、布林带下轨支撑; MSFT 出现 MACD 背离,需要注意短期回调; TSLA 出现均线死叉,提示中期下行压力。 这就不是简单告诉我“今天涨了多少”,而是帮我回答一个更交易化的问题: 这只股票现在为什么值得关注,关注点在哪里? 3. 个股详情页更像交易前检查清单 普通网站的个股页往往是行情 + 新闻 + 财务数据。 我这个站的个股详情页更偏“交易前判断”:价格、评级、52 周区间、K 线/成交量、基本面、技术信号、资讯都放在一起。 它的价值是把“发现信号”之后的下一步接住。 扫描器告诉我哪只票有异动,详情页帮我进一步看这个信号有没有质量。 4. 提醒中心是我最喜欢的功能 这个站不是看完就结束,而是有 Alerts Center。 里面可以看到全部提醒、已触发提醒、等待中提醒,还能新建价格/信号提醒。 比如: NVDA 突破 900 美元触发提醒; AAPL 出现 MACD 金叉提醒; TSLA 等待成交量异常触发; META 跌破支撑提醒止损风险; AMZN 突破近期高点提醒动能增强。 这和普通美股网站最大的差异就在这里: 它不只是信息展示,而是把“观察 →

Chat-based AI investigator that analyzes crypto transactions, profiles wallets, and traces stolen funds.
loopedcircuits · HN
AI Crypto Investigations/Research Agents

A global marketplace where users discover creators and purchase digital products with secure delivery.
@quatcrypto · X

Find price gaps between Polymarket and Kalshi prediction markets to identify arbitrage opportunities.
@b1llymane · X
startup #1 of 12 this year: it catches sportsbooks pricing the same game differently, bet both sides, profit either way. mostly built between 11pm and 3am, which felt thematically correct

Calculate expected value of gacha card packs using real pulling data.
@wizz_spellman · X
C级卡确实有点让人难受 这也是我们做这个黑客松项目的初衷之一 我们做了一个工具 基于真实的抽卡数据而不是官方公布的期望值 算出卡包真正的期望价值 让大家抽卡更有数 体验地址 小提醒 目前模型还比较年轻 数据样本还在积累中 随着时间推移和数据增多 预测会越来越准确

Test and benchmark AI trading agents against historical market data.
remote_ctrl · HN
BotTrade – a replayable benchmark for autonomous trading agents
View Bitcoin bottom and top-risk signals from composite technical, on-chain, and sentiment indicators.
@BDNcontent · X
Every post in my LinkedIn timeline is about one of two things right now: 1) Github is the new necessity for marketers, and 2) Substack integrated Pangram to "combat AI slop." The first one is kind of cool, I made my own Github this week to publish the first tool I vibe coded on my own. The second one is kind of annoying because while some people already know AI detectors are almost useless, some are also hailing Substack as being forward-thinking and righteous for bringing down the big bad AI sloppers. What Substack did boils down to a marketing ploy. The hot trend in advertising/marketing right now is being anti-AI because a lot of people are sick of AI slop, understandably. But Pangram isn't a fix for anything other than the most generic one-prompt outputs, i.e., someone goes to ChatGPT and says "write me a long-form newsletter," and they copy + paste the output. That will obviously be flagged. If someone has a custom project or GPT they trained extensively,

Scan smart contracts for vulnerabilities and get gas optimization suggestions.
@Ninjafromqueens · X
Hello! I am the founder and owner of Audit your smart contracts before deploying.

Trade crypto and forex with live markets, calculators, and a private trade journal.
@GoogleAiStwj9h · X
Check out what I just built with Lovable!