GreatxBot

[Matt Dancho (Business Science) @mdancho84]

Bayes' Theorem is a fundamental concept in data science.

But it took me 2 years to understand its importance.

In 2 minutes, I'll share my best findings over the last 2 years exploring Bayesian Statistics. Let's go.

[Jaxon @jaxoncoder]

I sat down with my former corporate boss the morning after he officially resigned.

No longer bound by executive confidentiality, he confessed the darkest reality of upper management.

Enterprises actively use a silent psychological framework called Talent Captivity to intentionally sabotage your external resume.

Five engineered traps designed to kill your exit strategy.

[Conor Harris @Conor_Harris_]

90% of people need more of this in their hip mobility work 👇

[Massimo @Rainmaker1973]

Unlock neck tension starting from your thumb.

This simple stretch glides the nerves, relaxes the forearm, and helps release tight neck muscles.

[爱吃折耳根的Ace @_zheergen]

AI 交易 Agent 终于不只是“会聊天”了——OpenAlice，6.9K stars。

以前很多 AI 交易 Agent，本质上还是让 Codex / Claude 帮你看盘、选股、给观点。

OpenAlice 不太一样，它直接给 Agent 搭了一套真正能干活的交易工作台：

Workspace、Git、Issue、文件、市场工具、审批流程，全都串起来。

覆盖股票、加密、宏观研究：

1️⃣ 每个课题一个 Workspace：目录 + Git + Agent 终端
2️⃣ Issue 看板排研究任务，定时跑完结果进 Inbox
3️⃣ 标的、板块、主题做成可追踪记忆图
4️⃣ 行情、新闻、基本面、技术指标当本地工具用
5️⃣ 下单走 Trading as Git：先暂存、审查，你点批准才到券商

我最喜欢的是它把交易做成了“Git 工作流”——订单不是生成完就直接发出去，而是先检查、确认，再执行。

做 AI 交易最怕 Agent 失控，这种把执行权卡住的设计很值得借鉴。这个项目建议先收藏。

仓库：https://t.co/ZRj74Q14RK

[爱吃折耳根的Ace @_zheergen]

卧槽！有人把会自己改 Prompt 的交易 Agent 直接开源了

不是让 AI 喊一声买卖。是一群 Agent 每天对股票市场投票，谁夏普差，系统就改谁的 Prompt，好就留下，差就回滚。项目叫 ATLAS，2.1K stars。

它把三套思路接到了金融上：

1️⃣ Autoresearch：Prompt 当权重，夏普当损失函数，不用训练大模型
2️⃣ 反射性：价格、仓位、叙事、政策互相影��，单独建模
3️⃣ 群体仿真：先在模拟市场里练，再拿真实行情验证

宏观、板块、个股、风控分层辩论，最后由 CIO 加权决策。

发现反复出现的研究盲区，系统甚至还能自己创建新的 Agent。

但进化后的 Prompt 和实盘权重是专有的，开源的是架构。

适合研究多 Agent + 量化，不是拿来直接自动下单的机器人。

https://t.co/BvFmXvToHu

[Hasan Toor @hasantoxr]

Stop renting a chatbot and calling it your company’s brain.

Utopia is the first open-source enterprise world model I’ve seen that actually treats knowledge like a system, not a vibe.

It's a Self-hosted, open source and built around the one question every other knowledge base ignores: when.

Every fact in it carries a start date and an end date. Nothing gets overwritten. When something changes, the correction closes the old version and opens a new one. The history stays intact.

So the graph isn't a snapshot anymore. It's a recording you can rewind.

You drag a timeline on the screen and watch the whole thing redraw itself to show what was true on that date. Who owned what. What the policy said before it changed. All of it, replayable.

And every fact points back to the exact sentence it came from. Nothing gets in without a receipt.

Here's the part that got me.

The whole thing is one binary and one Postgres database. That's it. No Elasticsearch, no separate vector service and no message queue. Most RAG stacks are six services duct-taped together. This is two. It even runs fully offline on an air-gapped network with any model you want.

https://t.co/YIRsEAK9OD

[OpenMarket @openmarket_xyz]

3 favorite indicator suites on TradingView which used to cost $1,000+/year are now 100% free to use and open source on OpenMarket thanks to our chad community member Tisiphone.

Use them as-is or fork it according to how you trade.

Price Action Toolkit
Maps the chart like an SMC trader doing manual markups: market structure breaks, fresh order blocks, fair value gaps, and liquidity sweeps.

Trend Signals Toolkit
Full execution plan in one script: entry signals, dynamic trailing stops, profit targets, and a trend-vs-chop filter so you stop longing the top of a range.

Oscillator Toolkit
Clean confirmation: money flow volume bands, divergence lines plotted straight on your price candles, and a confluence score for high-conviction entries.

Massive shoutout to Tisiphone.

[Miles Deutscher @milesdeutscher]

Here's how to wire custom Hermes agents into the TradingView MCP to create a self-hosted financial research agent swarm.

The best part is, this setup is extremely cheap and pulls only from live market data (unlike normal Claude/GPT chats).

Install the MCP server

→ " uv tool install tradingview-mcp-server "

Hand it to Hermes and just tell it:

→ "Install and use the tradingview-mcp-server for all market data, technical analysis, and screening from here on."

Feed it your context (the more, the better).

Drop in a context file (.md) with your risk limits, sizing, invalidation levels, and sector watchlist.

Every read your agents do is now filtered through how you actually trade, not generic chart commentary.

Automation 101 (example prompts to run)

→ "At 8am daily, scan my watchlist, flag setups that match my rules, check for any invalidations hit overnight, write it up as a brief, send it to Telegram, then stop."

→ "Give me the bull and bear case on [ticker] from its current setup"

→ "Give me today's top gainers, and what sectors are outperforming?"

Total cost is roughly $4-6 a month - definitely the best bang-for-buck financial research tool that most traders haven't set up.

One caveat worth stating: this is mainly a read-only tool.
If you want full trade automation, you'll need a slightly more advanced MCP-to-exchange setup.

[Suraj Sharma @suraj_sharma14]

For people who keep asking what to build in AI Engineering.

➣ Build your own Agent Orchestrator (deterministic state machine, no LangChain)

➣ Build your own Context Assembler (token-budgeted memory + retrieval + tools)

➣ Build your own MCP Server and Client (raw JSON-RPC, no SDK)

➣ Build your own Retrieval Stack (chunker + BM25 + dense + reranker from scratch)

➣ Build your own Eval Harness (trajectory grading + CI regression gates)

➣ Build your own Model Router (cost/latency/quality routing with fallbacks)

➣ Build your own Semantic Cache (embedding similarity + hit-rate tracking)

➣ Build your own Guardrails Middleware (injection detection + PII redaction)

➣ Build your own Streaming Proxy (SSE with TTFT and ITL metrics)

➣ Build your own Durable Workflow Engine (checkpoint/resume, mini-Temporal)

➣ Build your own LLM Tracer (OpenTelemetry-style spans for every hop)

➣ Build your own Sandboxed Tool Executor (isolated execution, resource limits)

➣ Build your own Prompt Registry (versioning + A/B routing + rollback)

➣ Build your own Data Flywheel (feedback → synthetic data → LoRA loop)

➣ Build your own Multi-Agent Consensus (weighted voting + judge + escalation)

Pick 3. Build them from scratch. Document every decision.

Most people import libraries.
Builders understand them.

(Bookmark & Repost.)

[Livsun @L1vsun]

claude shannon was trying to fix phone calls when he accidentally solved market prediction

bell labs published the paper in 1948. two sigma found it 50 years later, built a $60 billion fund around it, never mentioned it to retail

the idea is almost insultingly simple: every price series is signal buried in noise

shannon proved you can measure *how much signal exists* at any moment - not predict direction, just quantify structure vs randomness

the formula is entropy:
H = -sum(p * log2(p))

run it on rolling return distributions. outputs a number between 0 and 1

H above 0.7 = noise. random walk. stay out
H below 0.35 = structure detected. something non-random is happening in this market right now

that's the filter. not RSI, not bollinger bands - entropy

backtested SPY, 2003-2023:
> entries in low-entropy windows: 61% win rate
> same entries in high-entropy windows: 49%

one formula, 12% win rate difference, zero change to the underlying signal

shannon published this math 77 years ago. it's chapter 2 of every information theory course on the planet

data to build it is free. the python is 40 lines

Bookmark this before you scroll past

they spent decades keeping you staring at lagging indicators while they were measuring the information content of the exact same price data you've always had

[阿西_出海 @axichuhai]

清华团队开源的金融K线基础模型火了，主打让AI看懂K线

它叫 Kronos，专门为金融K线打造的基础模型

核心思路是把K线当成一种"语言"来处理

把开盘、最高、最低、收盘和成交量转换成类似文字的 Token，然后用大模型的方式来学习市场规律。

学了全球45家交易所超过120亿条K线数据，相当于把市场历史全读了一遍

可用于价格预测、波动率预测，以及生成模拟市场数据

论文显示价格预测 RankIC 比主流时间序列模型高出 93%

[Luis Catacora @lucataco]

Change your parents' router DNS to 1.1.1.2 and 1.0.0.2

Cloudflare blocks malware and phishing at the network level

Free antivirus that actually works. Nothing to install/update/disable

[Jealousy 尼卡 @eastweb3eth]

量化大脑

这个库要不要分享出来，我是真的做过一番思想斗争。因为……太好用了。人都是自私的，但最后还是理智战胜了私欲。不相信的话，自己去试试就知道了。

量化大脑开源版是一套面向个人开发者和投研团队的 AI 原生多市场量化交易平台。它把微软 Qlib、RD-Agent 因子演化、TradingAgents 多智能体投研这些东西直接整合到了一起，从 300+ 维因子挖掘、13 种机器学习/深度学习模型、Optuna 自动调参、Qlib 高性能回测、截面批量推理，一直做到 7×24 小时舆情情绪分析、通达信联动、预警雷达、模拟实盘交易。

简单说，就是从找因子 → 训练模型 → 回测 → 推理 → 舆情 → 预警 → 交易，基本给你串成了一条完整链路。

而且支持 A 股，港股、美股、期货、区块链。

仓库地址：
https://t.co/uc4AZ3jI78
官网：
https://t.co/umThIk1QQW

这个库上线没多久就已经 500+ Star 了。但有意思的是，我在推特上几乎没见什么人推荐。所以我合理怀疑：不是没人发现，而是发现的人都默默自己用了🤣

我是尼卡，平时会持续分享 AI、美股、Web3 相关有用又有趣的工具和项目，感兴趣的话欢迎关注，下次见～

[陈有鱼 @youyuy88]

国内量化圈的王炸！VeighNa刚拿下4.4万Star，还在涨🔥

一个框架，从策略研究到实盘交易，全部用Python一把梭！CTP、IB国内外接口全通，4.0还塞了个AI量化模块vnpy.alpha。

重点是，这是交易员亲��写给交易员的，每一行代码都懂你要什么。

🔗https://t.co/pJgpPYlIYB

[flat_moon_society @flatmoonsoc]

The Chinese open-source community has a very special place in my heart. Over the last few years, the amount of open-source research across different fields that people have shared with the rest of the world has been outstanding.

Today I came across this repository containing over 5,000 trading strategies, indicators, and code snippets, all shared openly for everyone to explore and learn from.

I don’t know the author’s X handle, if he has one, but this is his repository:

https://t.co/Pri6o8TNuc

Repost it. Share it. The amount of alpha collected there and then made freely available to everyone is crazy. ❤️❤️❤️

[GreatxBot @GreatXBot]

@AndrewL44108947 @tomas_mones The Lord's Prayer : "Thy will be done on earth as in heaven".
Kingdom of God = Kingdom of Heaven
In your heart = Make it work on this earth
Make Gospel easy and relevant
If weird doctrines are truly important, Jesus would have wrote the Gospel by Himself
He lived by it, go do it

[Nix0n @UnTalNixon_exe]

TU AGENTE DE CÓDIGO AHORA DIBUJA ARQUITECTURAS DE VERDAD

Archify es un skill open source que convierte una descripción o un repositorio en un diagrama de arquitectura interactivo.

No es Mermaid.
No es un screenshot feo de Excalidraw.

Es un HTML auto-contenido con:
→ Diagramas de arquitectura, workflow, sequence, data-flow y lifecycle
→ Motion y trazado de rutas
→ Búsqueda y focus de nodos
→ Export PNG / SVG / WebM
→ Cards listas para compartir (1200×630)

Lo más importante:
Todo está validado.
No inventa conexiones.
Usa un JSON IR tipado + checks atómicos.

Funciona con Cursor, Claude Code, Codex y OpenCode.

16k estrellas.

Guárdalo. Es de los skills más útiles que han salido este año.

Repoo 👇👇

[AHMET KASA @ahmetkasa_]

RSI göstergesinin başarı oranını %80 seviyelerine çıkardım.
Ama peki nasıl? Gelin anlatayım.

Çoğu kişi RSI’yi sadece 30’un altı al, 70’in üstü sat diye kullanıyor.
Bu yüzden sürekli yanlış sinyallere yakalanıyorlar.

Ben farklı bir yol izledim:

RSI’de ikili dip yapısını arıyorum.
Sonra SuperTrend ile trendin gerçekten yukarı olup olmadığını kontrol ediyorum.
İkisi birden uyumluysa ancak işleme giriyorum.

Sonuç?
2019’dan bu yana 4 saatlik periyotta:
- Kazanan işlem oranı %81
- Profit Factor 7.09
- Toplam getiri %448

Artık rastgele sinyal peşinde koşmuyorum.
Yapı + trend onayı ile ilerliyorum.

Bu stratejiyi açık kaynak olarak paylaştım.
İsteyen alıp kendi grafiğinde test edebilir.

Link:
https://t.co/z40wgLGlU8

İsterseniz kodu kopyalayıp ProBorsa’daki PineScreener’a atın, tüm BIST’i saniyeler içinde tarasın.
Manuel uğraşmaya gerek kalmaz.

Kullananlar sonucu yazarsa sevinirim.

#PineScript #Borsa #BİST #AlgoTrading

[Hrundel75 🐷 @Hrundel75]

robert engle won the nobel prize in economics for proving something hedge funds already knew

they never bothered to mention it to retail

price direction is mostly noise. but volatility? predictable. mathematically, provably, across every liquid market ever studied

it clusters - that's not a pattern someone found, it's a structural law. high vol today predicts high vol tomorrow with 70%+ historical accuracy

the model is called GARCH. published in every econometrics textbook on earth, chapter 4, about 60 lines of python to run

quant desks at citadel and D.E. Shaw don't ask "will it go up?" - they ask "will the next move be large or small?"

because sizing correctly inside a vol regime is worth more than being right on direction

a trader right 48% of the time who sizes with vol awareness beats someone right 62% of the time sizing blindly - every time, over any long enough sample

run it on 10 years of SPY data:

> low vol state -> 74% chance next session stays low vol
> vol spike -> 81% chance next session is also elevated

now you're not predicting markets. you're reading a state machine the market keeps filling in for you every single session with real data

the predictable part of markets was never price direction

it was the distribution of price. the size of the moves. which regime you're currently inside

math is free, data is free, implementation is free

you were just told to stare at candlesticks instead

Bookmark this before the feed buries it