🏆 Bitget Hackathon S1 赛道一参赛作品：XAU Session Breakout 黄金24小时波动但人需要睡觉，震荡行情频繁假突破，错过趋势行情的大幅收益。 因此我做了一个基于 Bitget Agent Hub 构建的 GetAgent Playbook，跟踪伦敦（07-13 UTC）和美盘（13-20 UTC）两大交易时段的高低点作为关键参考位。突破上个时段高点做多、低点做空，止损设日波动的 25%，止盈 2 倍盈亏比，趋势跟随保本损每盈利 1% 跟进 1%。 1年回测（1小时K线）：• 夏普比率 3.01 • 盈亏比 3.21 • 胜率 50% • 最大回撤 < 2% • 16笔交易全部在非美盘时段执行 已发布 GetAgent Playbook，真实可运行的 Demo，不是 PPT。 Demo: github.com/liuye884-BTC/xau-… #BitgetHackathon @Bitget_AI

Thrilled to present our final submission for the #BitgetHackathon S1! Meet WhalePulse AI - a Cross-Market Intelligence Engine built for precision. 🌐🧠 As shown on "image" our live dashboard tracks real-time prices and market sentiment. Today, it flagged an Extreme Fear zone (12/100), triggering our contrarian opportunity filters! 📊🔍 Here is how our engine breaks down the market across different layers... 👇 @GracyBitget @BitgetAI @Foresight_News #BuildInPublic #Crypto #Web3

GetAgent Playbook early access: I built one risk governor and ran it across three markets. It now holds the top 3 of the Studio leaderboard. Catalyst Discipline (SOL): score 59, 7.6%, Sharpe 2.56 ETH Discipline: score 52, Sharpe 2.33 BTC Discipline: score 52, Sharpe 2.31 Same thesis on all three: stay flat by default, only trade when trend, momentum, and volatility align. Few trades, high Sharpe, near-zero drawdown. Discipline over activity. Full feedback below. @Bitget_AI #BitgetHackathon

Who says hackathons are only for expert developers? Building my entry for the @Bitget_AI Hackathon S1 and it’s been an interesting journey so far. Currently working on a multi-agent trading engine that debates market scenarios, manages risk, and generates structured trade decisions. Excited to see what everyone is building. 🚀 Sneak peek of my trading agent 👀👇 #BitgetHackathon #BitgetAI

For the longest time, I've been the guy collecting ideas. Saving AI startup concepts, Testing new tools. Watching builders ship products. Always telling myself "I'll build something someday" Well, someday is now. I'm officially joining the @Bitget_AI Builder Base Camp Hackathon S1. As a social worker, builder, and active trader, I've always been interested in solving problems that make life easier for people. Whether it's helping individuals navigate challenges in everyday life or helping traders navigate complex markets, the goal is the same: turning confusion into clarity. This hackathon is my opportunity to move from simply using tools to building solutions. I believe the best technology isn't just useful for traders it's useful for people. That's the mindset I'm bringing into this journey as I explore ideas that can create real value, both in crypto and beyond. #BitgetHackathon

Meet APEX: our Track 1 submission for the #BitgetHackathon APEX reads live Bitget market data, weighs conflicting signals, and executes trades autonomously using a strict 10-step AI decision pipeline. No human in the middle. Just pure, adaptive reasoning 📈 @Bitget_AI

We just shipped a major risk management upgrade to Reflex Reflex now uses live 24h volatility data from @bitget to dynamically adjust stop-loss and take-profit levels before every trade. High volatility → Wider SL/TP ranges → More room to capture larger market moves Low volatility → Tighter SL/TP ranges → Faster exits and improved capital efficiency We also added full Qwen decision logging to the telemetry console, making every AI invocation visible in real time. Another step toward building a trading agent that adapts to market conditions instead of relying on fixed rules. Base Camp Hackathon S1 active cc: @Bitget_AI #BitgetHackathon #Bitget

Built Hellix8 for Bitget Hackathon Track 3 🟧 Problem: rToken traders hold static bags while macro sentiment keep shifting. What we built: a GetAgent Playbook that reads SPY regime fear/greed, ranks MSFT/AAPL/NVDA, rotates into top 2 weekly, with drawdown gates. 90-day sandbox: 2.11% return, 2.66% alpha vs equal-weight. Published on GetAgent Playbook (v0.0.4) ... real demo, not just slides. Demo: [loom.com/share/7f68abe07a054…] #BitgetHackathon @Bitget_AI

I'm staying fully within Bitget's ecosystem. Building on GetAgent and GetClaw. The platform has everything I need right now. My work is designing the decision logic on top of it and documenting the process publicly. #BitgetAI #BitgetHackathon

the @Bitget_AI hackathon gave me an idea to build my trading agent. you might be surprised that the brain isn't using a frontier model. check out the video of the timelapse for my trading agent called WYRM. as you can see, it's using Gemma-4-E2B. back to work, time to improve! registration of #BitgetHackathon still open. Huge thanks @googlegemma

在好哥们 @deepbluue 的帮助下，做了个参赛 Demo 哈哈哈哈哈哈哈。 xiaomaolu.github.io/signal-d… 根据市场情况，多个 Agent 协同讨论，制定交易策略的模拟。 当前显示主导为 FOMC 点阵图。风控建议，FOMC 风险足够近，风控台不会因为次级数据就贸然提高风险。 目前还在优化中。 报名链接：bitget.com/zh-CN/campaigns/d… #BitgetHackathon @Bitget_AI

The question that got me into this hackathon was this simple. I've been explaining agentic trading to people for months. Do I actually know how to use it myself? I'm joining Bitget Builder Base Camp Hackathon S1 to find out. I've spent meaningful time inside Bitget's AI ecosystem, using the tools, studying the products, covering the space as a creator. At some point that stops being enough. You want to actually build something with what you know. What draws me to AI x Crypto is the problem most traders don't talk about openly. Having access to powerful AI tools but no real system for using them. Everyone has a platform. Very few people have a framework. That gap is what I want to close. My plan is to enter the AI Stock Trading track and design a structured trading framework using Bitget's native AI tools. Document the logic. Document the process. Share everything publicly as I go. What I'm hoping to learn is how to translate trading intuition into something an AI agent can actually act on. That's the work. What would you build if you had one month to create an AI × Crypto product? I'm building in public. I'm starting now. Join me. Link: bitget.com/campaigns/d8a2a61… #BitgetAI #BitgetHackathon #SpaceX #SPCX

Tried GetAgent Playbook early access for my @Bitget_AI hackathon build. Imported Catalyst Discipline into Studio, a risk governed SOLUSDT strategy that stays flat by default. Cloud backtest, validated, published: 7.6% return 2.56 Sharpe -3.5% max drawdown 20 round trips #1 on the Studio leaderboard at the time Idea to validated Playbook to backtest evidence to public ranking, in one sitting. Backtests are evidence, not promises. The same strategy is paper running forward from here, and that ledger is the real test. #BitgetHackathon

这次体验了一下 Bitget GetAgent 的 Playbook，感觉它更像是这次黑客松里最值得关注的基础设施方向之一。（我的视频里有详细解说） Playbook 本质上不是一个简单的信号列表，而是一个标准化的策略运行单元。 它把一套交易策略需要表达的东西都结构化了： -策略逻辑 -适用行情 -风险边界 -交易标的 -可调参数 -运行频率 -信号输出 -是否支持自动交易 这意味着策略不再只是散落在本地脚本、Telegram 群或者黑箱 bot 里，而是可以被创建、发布、订阅、运行和管理。 首先，我这次先订阅了一个现成 Playbook：「智能捕捉均值回归窗口」。 我发现它不是传统网格，而是偏均值回归：过滤掉过低和过高波动，只在相对适中的市场环境里捕捉超买/超卖后的价格回归。 后来，我又用 GetAgent skill 创建了自己的 BTC/ETH Playbook。 逻辑来自本地已有策略：多周期趋势过滤 短周期突破触发。 第一版只支持 signal-only，第二版改成支持 follow trade，也就是在有信号时可以自动执行交易。 整个流程跑下来，GetAgent 的意义会变得更清晰： 它不是单纯做一个 AI 聊天入口，也不是再造一个 Telegram bot。 它更像是把“策略开发 → 策略封装 → 发布 → 订阅 → 自动运行 → 交易执行”这条链路标准化。 这对开发者和交易用户都很关键。 对开发者来说，Playbook 是策略产品化的容器。 你可以把本地策略打包成可被平台识别和运行的标准资产。 对用户来说，Playbook 降低了理解和使用策略的门槛。 用户不再只是看到一个买卖点，而是能看到策略适合什么行情、参数怎么改、风险在哪里、是否会自动交易。 当然，自动交易不是魔法。 策略仍然会遇到假突破、连续止损、极端行情。 但至少在 Playbook 框架下，策略的逻辑、权限和执行边界是清楚的。 我觉得这就是 GetAgent 黑客松项目最有意思的地方： 它不是简单地“让 AI 帮你交易”， 而是在尝试定义一套 AI Agent 时代的交易策略发布与执行标准。 如果说上一阶段的交易 bot 是脚本化和群组化的， 那 Playbook 更像是把策略带入了平台化、可组合、可订阅的阶段。#BitgetHackathon 对了还能报名，赶紧来，报名完去TG研究怎么拿奖～：bitget.com/zh-CN/campaigns/d… @Bitget_AI @Bitget_zh @GracyBitget @33Bitget

I am building TradePaddy for the @Bitget_AI Hackathon. TradePaddy is an AI trading OS that analyzes opportunities, tracks performance, remembers trades, scores risk, and helps traders improve. The goal isn't just better trades. It's becoming a better trader! #BitgetHackathon

I'm building the Narrative Memory Agent for @Bitget_Global AI Hackathon S1 🧠 Most trading agents are stateless. Mine remembers. It detects forming crypto narratives (AI coins, RWA, meme supercycles), queries a memory of historical outcomes to time entries, then writes back what actually happened, so the next cycle it knows more. Live dashboard: narrative-memory-agent.up.ra… Built with @Bitget_AI Agent Hub Skill Hub #BitgetHackathon #AIAgent #CryptoTrading

challenge accepted, @GracyBitget for the @Bitget_AI Hackathon, we're building an autonomous trading system that doesn't rely on a single indicator or a simple buy/sell bot the system continuously monitors multiple live market signals, feeds that data into an AI reasoning layer powered by Qwen, evaluates market conditions, weighs potential opportunities, and generates a conviction-based trading decision before execution. the goal is to create a workflow where market data, analysis, reasoning, and execution happen in one connected loop rather than requiring traders to jump between multiple tools still refining the intelligence layer, improving signal interpretation, and stress-testing decision making, but it's exciting to see the vision gradually become reality looking forward to competing alongside other builders in the #BitgetHackathon

最近 AI Agent 交易越来越热 但我一直觉得，真正有价值的不是 AI 会不会喊单 而是它能不能真的变成一套 可执行、可验证、可持续运行的交易系统 这也是Nana为什么会关注 @Bitget_AI Hackathon S1活动的原因 它的方向很明确： 不是泛泛地聊 AI 概念，而是围绕 AI × Crypto Trading，去做真正能落地的产品 这次主要有几个方向： 🔴交易 Agent 让 AI 自主看盘、分析行情、生成策略，甚至执行交易流程 🟢交易 Infra 为 Agent 提供更稳定的底层工具、数据、执行和风控支持 🟡美股 AI 交易 围绕美股代币化资产，用 AI 去构建新的交易策略和应用场景 我觉得这类 Hackathon 的意义不只是比赛本身 更重要的是，它会吸引真正懂交易、懂产品、懂开发的人，一起探索 AI Agent 在交易里的真实边界。 如果你是开发者， 或者你有自己的交易想法， 又或者你正在研究 AI Agent Crypto， 这次都可以去看看。 总奖池 50,000 USDT，全球线上参与 报名截止 6 月 14 日 感兴趣的话记得按时报名，别错过啦 报名链接： bitget.com/zh-CN/campaigns/d… #BitgetHackathon