freshcrate
Skin:/
Home > AI Agents > openclaw-autotrader

openclaw-autotrader

A 30-day public U.S. stock challenge: follow a 5000 HKD 🦞 claw through live market days.

Why this rank:Strong adoptionRecent releaseHealthy release cadence

Description

A 30-day public U.S. stock challenge: follow a 5000 HKD 🦞 claw through live market days.

README

Day 44 Poster

30-Day OpenClaw AutoTrader Challenge

Watch a 5000 HKD 🦞 claw take on 30 public market days in U.S. stocks. 看一只起步于 5000 HKD 的 🦞 claw,连续 30 天公开挑战美股市场。

Last synced by decision / 决策触发同步时间: 2026-04-22 04:36:39 CST

Why Follow This Repo / 为什么值得关注

  • a real 30-day live challenge, not backtest theater / 一个真实连续 30 天的实盘挑战,不是回测表演
  • public updates on decisions, recaps, and turning points / 决策变化、每日复盘和关键转折都会公开更新
  • a visible learning log that shows how the 🦞 claw updates its lessons over time / 一个公开学习日志,能看到 🦞 claw 如何随着挑战推进不断更新经验

Challenge Dashboard / 首页进度看板

Metric Value
Day / 当前天数 44 / 30 (146.7%)
Starting capital / 起始资金 5000 HKD
Current equity / 当前权益 HKD 4,822.77
Net PnL / 累计盈亏 -HKD 177.23
Open positions / 当前持仓标的 0 open: No live positions / 暂无公开持仓
Latest move / 最新动作 [US] HOLD / [US] 观望

30-Day Tracker / 30 天挑战总览

Learning Log / 学习日志

Follow how the 🦞 claw turns finished trades, missed timing, and quiet sessions into reusable lessons. 看这只 🦞 claw 如何把已完成交易、时机判断和观望时段,沉淀成可复用的公开经验。

Latest Snapshot / 最新概览

  • Updated / 更新时间: 2026-04-22 04:36:39 CST
  • Current book / 当前组合: No live positions / 暂无公开持仓
  • Floating PnL / 当前浮动盈亏: +HKD 95.89
  • Latest decision / 最新决策: [US] HOLD / [US] 观望
  • Next milestone / 下一阶段: Day 30 of 30
  • Public monitor / 公开监控: docs/public-monitor/2026/2026-04-22.md
  • Daily report / 每日报告: docs/daily-reports/2026/2026-04-22.md

Today's Trading Rules & Adjustments / 今日交易规则与策略调整

  • Execution objective / 执行目标: deploy pocket capital only when the expected edge remains meaningfully above fees and sizing limits, with no leverage and no shorting. 仅在预期优势明显高于手续费且满足仓位上限时动用口袋资金,不加杠杆、不做空。
  • Session discipline / 时段纪律: live decisions stay inside regular sessions, capped at 5 trade(s) per hour, with a 8% cash reserve and HKD 5000 daily loss stop. 实盘决策仅在常规交易时段内执行,每小时最多 5 笔,并保留 8% 现金缓冲,单日亏损达到 HKD 5000 即停止扩张。
  • Live pools today / 今日实盘池: US: AAPL, MSFT, META, GOOGL, AMZN, NVDA, AVGO, MU, BABA, RIVN, AMD, QCOM, TSM, TSLA, ORCL, WMT, LLY, JPM, XOM, V, MA, ASML, JNJ, ABBV, PG, BAC, HD, COST | HK: 0388.HK, 1810.HK, 1024.HK, 1211.HK, 3750.HK, 0700.HK, 9988.HK, 3690.HK, 9999.HK, 9618.HK, 9888.HK. 今日实盘池如上,按市场分别执行。
  • Observation focus today / 今日观察重点: themes 半导体, AI芯片, 云软件, 中概, 电动车, 手机链, 消费电子, CPO, 光模块, 存储; public observation pool US: none / 暂无 | HK: none / 暂无. 今日观察主题为 半导体, AI芯片, 云软件, 中概, 电动车, 手机链, 消费电子, CPO, 光模块, 存储,并同步公开观察池变化。
  • Explicit exclusions / 明确排除: 智谱, MiniMax, 三星电子, SK 海力士, 7709.HK stay out of the live universe when they violate the rules. 凡与规则冲突的标的(如上)均不进入实盘池。
  • Latest gate result / 最新门槛结论: No US candidate cleared the live entry bar. The strongest name, MU, still showed score -2.17, post-fee EV -1.53%, and win probability 59.3%. / 研究链路未稳定返回,但当前最高候选 MU 的费后 EV 为 -1.53% ,仍低于 live 开仓门槛 0.26% ,按小账户费后纪律继续 HOLD。

Latest Decision Basis / 最新决策依据

  • Result / 结果: [US] HOLD / [US] 观望
  • Rationale / 理由: No US candidate cleared the live entry bar. The strongest name, MU, still showed score -2.17, post-fee EV -1.53%, and win probability 59.3%. / 研究链路未稳定返回,但当前最高候选 MU 的费后 EV 为 -1.53% ,仍低于 live 开仓门槛 0.26% ,按小账户费后纪律继续 HOLD。
  • Decision basis / 决策依据: Regime: mixed tape; Path: standard decision flow; Model: research gpt-5.4, compare A; Purpose: standard review; confidence 0.22. / 市场状态:混合状态;决策链路:常规决策链路;模型:研究模型 gpt-5.4,候选比较 A;目的:常规审查;置信度 0.22。
  • Candidate check / 候选检查: Reviewed 5 active candidate(s). Top checks: MU (semiconductor) | score -2.17 | post-fee EV -1.53% | win 59.3%; ORCL (cloud software) | score -2.98 | post-fee EV -0.76% | win 52.8%; QCOM (semiconductor) | score -3.14 | post-fee EV -1.82% | win 48.9%. / 共检查 5 只活跃候选。靠前检查结果:MU(半导体) | 评分 -2.17 | 扣费后 EV -1.53% | 胜率 59.3%;ORCL(云软件) | 评分 -2.98 | 扣费后 EV -0.76% | 胜率 52.8%;QCOM(半导体) | 评分 -3.14 | 扣费后 EV -1.82% | 胜率 48.9%。
  • Watch next / 下一步观察: Wait for at least one active candidate to turn fee-adjusted expectancy positive and clear the live score buffer. / 等待至少一只活跃候选的扣费后预期收益转正,并越过实盘评分缓冲区。

Core Rules / 基本规则

  • Starting pocket capital / 起始口袋资金: 5000 HKD
  • Default market / 默认市场: US equities first, with HK monitoring when relevant / 以 US 市场为主,必要时监控港股
  • Public operation day 1 / 公开运行首日: 2026-03-10
  • Guardrails / 约束: whitelist-only, bounded deployment, no leverage, no short / 白名单、有限资金、不加杠杆、不做空
  • Disclosure boundary / 披露边界: publish strategy, holdings status, decision status, and daily activity only / 只披露策略、持仓状态、决策状态和每日交易活动

What This Repo Publishes / 这个仓库公开什么

  • current holdings with quantity / 当前持仓与数量
  • latest trade timing and execution rationale / 最新交易时机与执行理由
  • latest no-trade reason and next watch item / 最新观望理由与下一步观察点
  • public operating rules / 对外可披露的操作规则

Release History

VersionChangesUrgencyDate
main@2026-06-01Latest activity on main branchHigh6/1/2026
0.0.0No release found — using repo HEADHigh4/9/2026

Dependencies & License Audit

Loading dependencies...

Similar Packages

memindSelf-evolving cognitive memory and context engine for AI agents in Java. Empowering 24/7 proactive agents like OpenClaw with understanding and SOTA performance.main@2026-06-05
ai-trading-agent🤖 Build and optimize an AI trading agent using JavaScript for efficient and intelligent trading decisions on financial markets.main@2026-06-05
opencode-telegram-botOpenCode mobile client via Telegram: run and monitor AI coding tasks from your phone while everything runs locally on your machine. Scheduled tasks support. Can be used as lightweight OpenClaw alternav0.21.1
openclaw-model-bridgeConnect any LLM to OpenClaw — production-tested middleware for Qwen3-235B and beyondmain@2026-06-03
clawmetrySee your agent think. Real-time observability dashboard for OpenClaw AI agents.v0.12.386

More in AI Agents

@blockrun/franklinFranklin — The AI agent with a wallet. Spends USDC autonomously to get real work done. Pay per action, no subscriptions.
hermes-agentThe agent that grows with you
awesome-copilotCommunity-contributed instructions, agents, skills, and configurations to help you make the most of GitHub Copilot.
e2bE2B SDK that give agents cloud environments