Gerald @TheRealGCabrera

this is the most BROKEN github repo in trading right now real trading firms pay analysts, researchers, traders and risk managers millions a year - someone replaced the whole floor with AI agents that argue with each other before every single trade, and put it on GitHub for 100% FREE it's called TradingAgents. the paper is on arXiv, the repo has 112k stars and it hit #1 on GitHub trending first understand how it works: > 4 AI analysts read the market: fundamentals, news, sentiment, technicals > a BULL agent and a BEAR agent debate the trade > a trader agent turns the debate into a decision > a risk team (aggressive / neutral / conservative) attacks that decision > a portfolio manager approves it or kills it > it's a full trading firm. it just runs on your laptop here is how to run it yourself for FREE: 1. open Claude Code and ask it to clone github.com/TauricResearch… and install everything 2. add your Anthropic API key. Opus 5.5 as the deep-thinking model, Sonnet 5.5 as the fast one 3. type tradingagents in the terminal and pick a ticker 4. watch the 4 analysts write their reports in real time 5. read the bull vs bear debate. this is where the edge is - you see both sides before any money moves 6. check the risk team's vote. if the conservative agent flags it, ask yourself why 7. let the portfolio manager approve or reject, then plot every call on your chart 8. backtest it on past dates and paper trade it for a few weeks before you trust a single call i won't say anything more:

Сarm1ne @carm1nee

3 months ago

x.com/i/article/2085…

72 162 837 8.1M 3K

THIS IS A PURE F*CKING GOLD beyond SpaceXAI’s 146 official templates for Grok Bot, there are another 624+ ready-made bots you can use completely FREE. and almost nobody knows they exist full teams with specialist roles, bots that hand work to each other, separate verifiers that check the output, and entire multi-bot trading desks. all open source 20 GitHub projects worth opening first: 01 awesome-grokbot → a searchable index of public Grok Bot configurations, organized by job. find an existing setup instead of writing every instruction yourself github.com/kydlikebtc/awe… 02 botteams → pick a company team and copy its installer prompt. it creates named bots, a group chat, and proposes routines for your approval github.com/ellelion/botte… 03 rosterroom → 82 paste-ready team setups covering founders, operations, security, finance, and research. each includes roles, a first assignment, and approval boundaries github.com/codejunkie99/r… 04 grok-bot-profiles → Spec Writer, Bug Reproducer, PR Producer, and PR Verifier. the verifier works in a separate conversation so the builder doesn’t approve its own work github.com/HAEGONG/grok-b… 05 hypergrok-trading-desk → 7 specialist bots and 17 skills for market research, strategy testing, risk, execution, and review. starts with public data; trades require your approval github.com/galleonlabs/hy… 06 root-agent-skill-framework → a coordinator that sizes the team your request needs, creates specialists, and delegates daily management to leads github.com/MrBekoX/root-a… 07 awesome-grokbot by mergisi → paste START.md into a setup bot and choose an engineering, sales, marketing, success, operations, or personal team github.com/mergisi/awesom… 08 grok-bot-templates → 49 templates, 10 teams, 47 skills, and 8 routines. each setup defines the job, deliverable, boundaries, and first task github.com/cobusgreyling/… 09 botskills → paste-ready BOT.md setups with required integrations and a clear boundary: the action each bot must never take without you github.com/PramodDutta/bo… 10 grok-skills → 195 reusable playbooks for code review, debugging, research, writing, planning, and shipping. add a proven process instead of another vague instruction github.com/jaskirat1616/g… 11 crew-contract → a working agreement for bot teams: defined missions, named owners, evidence in handoffs, completion checks, and stopping when the budget runs out github.com/lsj210001/crew… 12 grokbot-for-gtm → an outbound playbook that connects prospect research, email, LinkedIn, replies, and meeting tracking. bring your accounts and approve the sends github.com/bcharleson/gro… 13 gojiberryai-sales-os → 13 sales roles covering buying signals, prospect research, enrichment, copy, follow-ups, qualification, and pipeline analysis through the GojiberryAI MCP github.com/romangojiberry… 14 grok-research → an equity research team with persistent researchers, scout reports, reviewed views, and a shared research database. research stays separate from live trading github.com/jaredtrichard/… 15 grok-bot-token-saver → a chief-and-workers setup that replaces dragging old conversations forward with short role cards and handoffs for the next session github.com/Chakhdz/grok-b… 16 grok-bot-cli → create bots and groups, send messages, and inspect threads from scripts. it reuses your signed-in desktop session without manual token copying github.com/ScriptedAlchem… 17 grok-bot-skill → let a coding agent list, message, and create Grok Bot teammates, connecting development work with the bots already running your workflows github.com/adamanz/grok-b… 18 awesome-grok-bot by majiayu000 → another searchable catalog of public bot shares, plus team recipes, setup guides, and community tools organized by job github.com/majiayu000/awe… 19 akeru-bot → a separate, self-hosted alternative with named teammates, per-bot tools, instructions, and memory. keep profiles and conversation history on infrastructure you control github.com/opencoredev/ak… 20 guaca → a separate desktop workspace where agents share conversations, delegate tasks, and use MCP tools. run the workspace locally or on your own server github.com/madebywelch/gu… start with one job you already understand. read the setup, connect the tools it needs, and inspect the first result before scheduling it the templates and source code are free. running them still uses your plan, APIs, or hosting then make the team share findings instead of repeating research. keep handoffs short, preserve completed work, and give failed operations a stopping point save this. your next useful bot might already be built the article below breaks down 7 workflow changes to make that setup cheaper and get more finished work from the same budget ↓

Gresham @Greshamtrade

4 days ago

x.com/i/article/2112…

34 21 193 254K 599

Bro just gave every entrepreneur a cybersecurity degree in one post.

EL GANADOR DEL HACKATHON OFICIAL DE ANTHROPIC HA PUBLICADO SU SETUP DE CLAUDE CODE Y ES ORO PURO 68 subagentes, 286 skills, 94 comandos, licencia MIT. ECC convierte Claude Code en un equipo de ingeniería completo. El plan va antes que el código. El test que falla va antes que el arreglo. Cada cambio lo revisa un contexto que no ha visto cómo se escribió. → Planificación: le das una frase y apruebas el plan antes de que exista una sola línea
→ Revisión: un revisor distinto por lenguaje que lee tu diff desde cero
→ Arreglo de builds: un corrector por toolchain, incluidos PyTorch y CUDA
→ Seguridad: barrido OWASP completo y escaneo de inyecciones en la config del agente
→ Arquitectura: detecta errores de diseño cuando aún son baratos, antes de la migración
→ Dominio: bases de datos, pipelines de ML, tests E2E y documentación La parte de seguridad es la que casi nadie configura. Una auditoría OWASP externa cuesta cuatro cifras y una semana. Esta la tienes lista antes de comer. Las skills cubren testing, Python, Go, Rust, Django, Next.js, Docker y CI/CD. Eso sí, no actives las 286 de golpe o acabarás con una config peor que la que tenías. Así se integra bien: → Haz un fork y léete el README antes de copiar nada
→ Empieza solo con el agente de planificación y el pack de reglas de tu lenguaje principal
→ Copia a tu carpeta .claude únicamente los agentes y skills que vayas a usar esta semana
→ Añade después el revisor de código y el de seguridad, que son los que más rinden
→ Pruébalo primero en un repo pequeño, no en producción
→ Si algo sobra, recórtalo. Si echas algo en falta, añade una sola pieza más La regla es simple: si no sabes para qué sirve un agente, todavía no lo necesitas. Quien lo integre bien este fin de semana se pasará el próximo trimestre revisando trabajo en vez de escribiéndolo. Guárdate este post, me lo agradecerás 🔖

angel @angeldot_

2 weeks ago

x.com/i/article/2109…

20 141 1K 1.3M 4K

there is no catch and that scares me Elon Musk posted an idea that hit 27 million views: "stop building startups. buy boring internet real estate and let bots run it" that's literally what Grok Bot does now. the whole setup costs $0 on top of the sub: → create a bot, name it Mogul → tell it to find one boring niche people search every day (it reads X and reddit complaints in real time, it's grok) → it builds a small directory site on its own computer and ships one useful page every night → it negotiates affiliate and referral deals itself, by email, in your name → publishing runs on a routine. telegram sends you the money reports → month one: traffic. month three: the first $217. it never stops after that your laptop stays shut. the site grows anyway every page is a tiny employee that works forever. it compounds like rent some of them run 3 sites each and haven't opened a single one. they don't post about it, and now you know why every month you wait, someone's bot ships another 30 pages into the niche you would've picked full playbook below bookmark this before someone takes your niche

Validate your skills in agentic AI workflows and AI-powered apps. Pass the online exam to earn a Credly badge, an official certificate and access to free credits. Limited-time offer: $1 instead of $100. Explore the certification and take the next step.

This is completely f*cking wild. A Stanford AI researchers gave Claude Code a second brain with JEV. The result? 80+ billion data points processed every 14 minutes. here's how it works: Jev stands at the door. every row has to get past it first. the easy ones it waves through on the spot. only the ones it can't call get sent upstairs to Claude. the old way: everything -> Claude -> big bill the new way: everything -> Jev -> the tricky few -> Claude Jev also tells you how sure it is on every call. that number decides who goes upstairs. just don't pick it out of thin air. test it first. the 20-minute setup: 1. take 200 real rows from your data 2. run them through Jev AND Claude 3. sort Jev's answers by how sure it was 4. find the point where Jev stops disagreeing with Claude. that's your cutoff 5. above it, Jev's answer is final. below it, the row goes to Claude then do the math. on a public benchmark Jev cost $0.015 per 1,000 rows and Claude cost $0.80. if Jev settles 80% of your rows: $0.015 + (20% × $0.80) = $0.175 per 1,000 rows instead of $0.80 that's ~78% off, with Claude still handling every hard case. 3 rules that keep it working: -> ask all your questions in one Jev call. topic, urgency and sentiment together, not three separate requests -> once a week, send 50 of Jev's "confident" answers to Claude anyway. if they start disagreeing, raise the cutoff -> never let Jev make the final call on anything you can't undo. refunds, bans, deletions go to Claude or a human where this pays off right away: -> support tickets: route by topic and urgency -> leads: hot, warm or skip -> comments: fine, spam or needs a look -> spreadsheets: fix category labels across thousands of rows Claude thinks. Jev picks. code ships.

kiosa @thegreatest_sv

4 weeks ago

x.com/i/article/2102…

23 79 455 1.2M 1K

Claude Code 刚刚通过逆向工程，在 90 秒内破解了我酒店房间的门锁。2014 年的加密技术。右上角的烟雾探测器也使用同一套协议——那个红色小灯不是电池指示灯。我启动扫描时，它闪了两次。 我现在躺在一个每晚 85 美元的酒店房间里。我打开 Claude Code，输入：「扫描范围内的每一台蓝牙和无线设备。告诉我哪些设备在与哪些设备通信。」 它找到了 19 台无线设备。其中 18 台是正常的。 设备 19：一个 BLE 控制器，使用着自 2014 年以来就没有修复过的固件。它控制着电子门锁、烟雾探测器、恒温器和房间里的保险箱。四台设备共用一个控制器。同一种协议。10 年没有任何更新。 我让 Claude Code 分析这套协议。整个过程用了 90 秒。 门锁使用一个 4 字节的滚动密钥，没有会话令牌。Claude Code 在第三次尝试时生成了一个有效的主密钥。我躺在床上，听见房间的门锁咔哒一声打开了。 接着，它发现烟雾探测器也连接在同一条总线上。照片里的红色小灯？那是控制器状态指示灯。它并不监测烟雾。它每隔 8 秒就会确认一次与门锁系统之间的 BLE 握手。 这意味着，任何处于蓝牙覆盖范围内的人——走廊、隔壁房间、停车场——只需要一条提示词和 90 秒，就可以： 打开你的房门 关闭房间内的警报 打开你的保险箱 而烟雾探测器会通过那个小小的红色闪烁，确认连接状态 检查只需要 2 分钟： 步骤 1 → 打开 Claude Code → 粘贴第一条评论中的提示词 步骤 2 → 它会扫描覆盖范围内的每一台 BLE 设备，并标记存在未修复固件的设备 步骤 3 → 如果发现问题 → 要求更换房间 → 要求换到配有机械门闩的房间 → 然后使用那个房间 每一家使用 2018 年之前生产的 ASSA ABLOY 或 Onity 门锁的酒店连锁都存在这个问题。这大约占全球酒店的 30%。 我本周测试了 3 家酒店。3 家全部存在漏洞。 看看照片。右上角。那个每隔 8 秒闪烁一次的红点。那就是握手信号。 保存这条提示词。下次入住酒店时 → 先运行它，再打开行李。

make claude @make_claude

3 weeks ago

x.com/i/article/2105…

0 14 89 454K 239

this is f*cking DANGEROUS thing on internet today one developer built a real-time global intelligence dashboard that tracks every world event 24/7 giving you the FASTEST market data to place trades and released it for 100% FREE i connected it to my Grok Bot & now my AI trading system sees wars, sanctions, oil shocks the moment they happen, long before they hit the news on TV or your timeline my agent sees the news data, maps it with the stocks that would be affected by the news and it trades them automatically 24/7 in trading, the person who sees it first makes real money, everyone else is exit liquidity here is how to build this AI trading setup, NO CODING NEEDED: 1. open worldmonitor app in your browser, switch to the finance version and spend at least 10 minutes exploring the globe and markets panel 2. get your World Monitor API key, this is what lets your AI read the dashboard for you 3. open Grok Bot and type “add a custom MCP server for World Monitor” with your key, it connects in seconds 4. sign up for a FREE Alpaca paper trading account, it gives you fake money to trade with, zero risk 5. copy your Alpaca API keys and ask Grok Bot to “add the Alpaca MCP server in paper mode” with them 6. tell Grok Bot in plain English: “every few minutes, check World Monitor for big world events and tell me which markets they could move” 7. give it your rules, like “if oil supply risk spikes, watch energy stocks, if a country’s instability jumps, watch gold” 8. best way is to set up one separate agent to map which stocks each piece of news would affect (i did the same) 9. before every trade, ask it to check what traders on X are saying and whether prediction market (Polymarket & Kalshi) odds have already moved, if the crowd hasn’t priced it in, that’s your EDGE 10. let Grok Bot set the entry, exit and stop for each trade, keep every position small and place it on your Alpaca paper account 11. ask it to message you each time with the exact world event that triggered the trade, then review every trade for a few weeks before you ever use real money if you want to build an AI trading system from scratch, just start with this article below:

Roan @RohOnChain

2 weeks ago

if you want to build an AI trading system from scratch, start with this article you will end up with real market data, real backtests and a bot that trades for you 24/7 x.com/i/article/2108…

87 143 993 1.2M 4K

Realmente no entiendo por qué nadie está usando esta configuración graphify, una herramienta que convierte cualquier carpeta en un grafo de conocimiento que tu agente puede consultar, acaba de superar los 126k estrellas en GitHub y entró en Y Combinator. La gente dejó de hacerle preguntas a la IA. Le están dando un escritorio, una memoria y un trabajo. Le das una carpeta a Claude Code, conectas tu correo, calendario y un navegador, y se convierte en un agente. Lee, recuerda, actúa y se revisa a sí mismo antes de hacer algo estúpido. No importa si eres fundador, marketer o estudiante, lo necesitas La configuración completa: > Crea una carpeta llamada AGENT, abre Claude Code dentro de ella y agrega un CLAUDE.md con una regla: "guarda cada decisión, persona y proyecto en la memoria/" > Ejecuta npm install -g @googleworkspace/cli, luego gws auth setup. Ahora lee tu Gmail, Calendario y Drive > Ejecuta claude mcp add playwright npx @playwright/mcp@latest. Consigue su propio navegador para abrir sitios, llenar formularios y extraer datos para ti > Ejecuta uv tool install graphifyy, luego graphify install, luego escribe /graphify .. todo en la carpeta se convierte en un grafo de conocimiento que consulta en segundos en lugar de excavar archivo por archivo > Ejecuta npm i -g jev-guard, jev-guard key "...", jev-guard install claude. Un modelo diminuto llamado Jev puntúa cada acción antes de ejecutarla. rm -rf se bloquea, un correo envenenado se marca, por unos 4 centavos por 1.000 verificaciones > Escribe /schedule y dile "todas las mañanas a las 7:30, envíame lo que entró, lo que redactaste y lo que espera mi sí". De ahora en adelante trabaja en una rutina 15 minutos para configurar, y nunca más empiezas desde un chat en blanco. repo: -> github.com/Graphify-Labs/ guárdalo y sígueme para más → @karlarboledas

This is f*cking gold.. Chinese researchers just published a paper that kills the hottest skill of the year: "The End of Context Engineering". The argument fits in one line: a context window packed by a human is a guess made before the model has looked. You pack what you would have needed. The model needs something else. Too much of the obvious, none of the surprising. So the paper hands the model an index instead: one line per thing that exists, never the content. The model names what it needs, pulls only that, runs the task, and every item gets credited or blamed for the result. What did not move the score gets dropped. What did gets pinned for the next task. Across 300 tasks: score 0.14 higher than the hand-built pipeline with a window 81% smaller. Three hundred prompts cannot be packed by hand. They can be scored by a loop. So you stop being the person who decides what the model sees. You become the person who names what exists and reads the score. Pipelines turn disposable. The index turns into the job. That is the exact system I run and documented below: one index file, 300 Kimi K3 tasks, 212 items pinned, 4,970 dropped, zero prompts packed by me. The whole paper in one line: index the shelf, not the page ↓

slash1s @slash1sol

2 months ago

x.com/i/article/2092…

26 16 194 675K 424

THIS IS A F*CKING GOLDMINE SpaceXAI’s Grok Bot team shared their bot templates for building a $1M one-person company in the marketplace you set the priorities, connect your tools, and review the work while each bot handles a specific part of building, marketing, selling, or running the business here are 10 Grok Bots i’d split across 4 departments: product & development: 1/ PM bot: turns your idea into a scoped brief with what to build, what to leave out, and how to know it’s finished 2/ Proto Bot: builds a working prototype inside your app and brings back a draft PR with screenshots so you can review the result 3/ Startup QA Bot: walks through your product in a test account, catches broken flows, and reports bugs with screenshots and reproduction steps marketing: 4/ Distro Bot: prepares campaigns, creative, and channel analysis, with your approval before publishing, sending, or changing spend 5/ AI Search Visibility: checks whether Google and AI assistants recommend your business, then shows which competitors appear instead sales: 6/ Outbound Prospecting: researches potential customers against your ideal buyer profile and drafts a relevant first message for each 7/ GTM Loop Closer: finds the customer promises and follow-ups buried in meetings, email, and Slack, then prepares the next action 8/ Sales Call Coach: reviews your call transcript or recording and gives you specific changes to make before the next sales conversation operations: 9/ Haggle Bot: checks your software bills for wasted spend, compares alternatives, and drafts vendor negotiations before renewals 10/ Alfred: maps responsibilities across your bots, spots duplicate jobs, and recommends the smallest useful setup for your business start with the department that keeps pulling you away from customers, connect its tools, and give one bot a real assignment save this and build your one-person company one useful role at a time all 10 are on the Grok Bot Marketplace ↓

9inches @0xnineinch

2 weeks ago

x.com/i/article/2108…

13 6 87 77K 204

HEDGE FUNDS BUILD THIS INTERNALLY AND CHARGE MILLIONS TO ACCESS IT someone built a full ML trading curriculum and put it on GitHub for free 27 chapters that takes you from raw market data all the way to a running strategy deep learning, causal ML, RL, autonomous agents, all with real working code 21k+ people already have it bookmarked

Ruuj @RuOnChain1

a week ago

The complete playbook for running a quant desk in 2026: tools, stack and the structure that keep you alive. x.com/i/article/2109…

11 21 172 170K 402

THIS IS PURE F*CKING GOLD Anthropic just dropped the ultimate autonomy loop for Claude Code The idea: run /goal in auto mode, and let the new Haiku 5.5 judge every turn True overnight autonomy with zero burn: Sonnet builds, Haiku evaluates at 340 tok/s, Opus signs off You just launch Claude Code like this: claude --model opus --dangerously-skip-permissions here's the split: → Opus 5.5 locks the goal condition and plans the overall architecture → Sonnet 5.5 writes the fixes, refactors modules, and runs test suites → Haiku 5.5 evaluates completion after every turn as an independent Stop hook → you stop burning Opus context on dozens of intermediate iteration turns Opus only shows up at three moments: → framing the goal condition before the loop starts → intervening if the loop hits the turn cap or an impossible verdict → signing off on the final git diff once the goal clears Opus frames · Sonnet builds · Haiku judges JEV engineering is the same move one layer down: the forks that need no thinker (which tool, path resolution, retry or abort) clear in milliseconds, and the big models only see the ones that split the full tree > Opus 5.5 runs the main session and defines the completion condition > auto mode approves safe tool calls so turns run unattended > Sonnet 5.5 patches code and runs test suites in an isolated copy > Haiku 5.5 evaluates transcript evidence: Met · Not yet met · Impossible > background check-in monitors idle tasks and backs off stuck loops > Opus 5.5 reviews the achieved transcript and signs off on the diff > JEV routes micro-forks in milliseconds paste the tree and this prompt into Claude Code ↓ "Rebuild my Claude Code setup around this tree: 1. Run the main session on Opus 5.5 with auto mode enabled 2. Route the goal evaluator hook to the new Haiku 5.5 via ANTHROPIC_DEFAULT_HAIKU_MODEL 3. Use Sonnet 5.5 for implementation and refactoring turns, keeping fixes isolated 4. Add one rule to CLAUDE.md: always bound autonomous /goal conditions with a verifiable end state and stop after 20 turns with no progress 5. Route fast deterministic forks through JEV Show me every change as a diff first, no edits until I say go" ↳ code.claude.com/docs/en/goal save it, it'll come in handy

mirku @mirku21

a month ago

x.com/i/article/2098…

29 78 678 2.1M 2K

It is becoming f*cking dangerous: someone published the full prompt for building trading agents. You paste it into Claude Code and it builds the bot while you have lunch No code. Just Ctrl+C / Ctrl+V. One prompt sets up the whole thing: Time: 1 minute Action: copy and paste into Claude Code The main problem with 99% of bots: one strategy for everything. But the market on a calm day and during a crash are two completely different worlds Quant mode: detects the regime (sideways, jittery, dumping) and switches tactics on the fly E-brake: a hard 15% drawdown limit and a risk check before every single order Hardcore backtest: replays the strategy on historical data with fees and slippage, and shows you exactly where it falls apart. a curve that only looks good on the past is the easiest thing in the world to build Nightly analyst: while you sleep it reviews every fill, writes up what went wrong, and hands you the notes in the morning Near-zero code: you paste the prompt and your parameters, Claude writes the engine, and you read the risk layer before anything touches a live account This used to take a whole development team and months of testing. Now it takes a single keyboard shortcut The prompt is below. paper trade it first, then save it

Hanako @hanakoxbt

a week ago

REA is the most dangerous repo on GitHub right now It gives your AI agent a real disassembler, so it reads any app down to the binary Every claim comes back with the address behind it I broke down the exact 8 steps: x.com/i/article/2110…

81 95 782 1.5M 2K

This guy shows you in 22 minutes how to learn to create videos automatically with Claude Code

this is the most DANGEROUS thing on internet today REA (Reverse Engineer Anything) - this open source MCP lets your AI agent reverse engineer any app, down to the binary i just reverse engineered TradingView, now I am using their premium indicators for FREE - software is cooked here is how to use REA to reverse engineer TradingView: 1. point REA at TradingView's publicly observable interface and map every major feature, from chart interactions to indicators and alerts 2. trace how the interface behaves when you switch symbols, change timeframes, add indicators, draw levels, and update chart layouts 3. use REA's available MCP tools to investigate authorized software artifacts and trace application logic wherever the evidence allows 4. feed those findings into Claude Opus 5.5, which turns the investigation into a technical specification and an implementation plan 5. rebuild the charting engine independently, with real market data, candlesticks, technical indicators, drawing tools, and persistent watchlists 6. add a quant backtesting engine that tests strategies against historical data, accounting for fees, slippage, and drawdowns 7. test every component, compare its behavior against documented and observable results, fix failures, and keep iterating until the system actually works 8. you end up building your own TradingView-style quant research terminal from scratch, with source code you control and features you can keep extending i won’t say anything more:

Roan @RohOnChain

a month ago

Jev is the FASTEST AI model ever built for trading It makes calibrated buy/sell decisions in under 100 ms That is one real decision on every single block, 24/7 In this article I've shown EXACTLY how to build HFT trading system with Jev (from scratch) x.com/i/article/2100…

224 412 3K 5.6M 12K

i went through every uncensored and abliterated models on hugging face and picked the best ones for each vram tier from 8 to 24gb. welcome to the danger zone ⚠️ ━━━━━━━━━━━━━━━━━━━━ 8gb · rtx 3060 ti, 3070, 4060, 5060, 4060 ti 8gb, 5060 ti 8gb · amd rx 6600, 6650 xt, 7600, 9060 xt 8gb · the base runs 42.8 tok/s on a 3060 ti 8gb out to a 96k window > 1. bonsai 2 27b heretic, 1-bit, 5.95gb, refusals 95 to 0 out of 100, 147k downloads this month huggingface.co/OS-Software/Te… > 2. bonsai 2 27b abliterated, refusals 83% to 0%, mmlu holds at 78%, 77k downloads huggingface.co/BoldingBuilds/… > 3. bonsai 2 27b crack, harmbench refusals 93% to 0% → huggingface.co/dealignai/Bons… older 8gb cards · gtx 1070, 1080 · the base runs 42 tok/s on a 2016 gtx 1080 > 1. gemma 4 e4b aggressive, 5.3gb, 0 refusals out of 465, 1.43m downloads this month huggingface.co/HauhauCS/Gemma… > 2. gemma 4 e4b heretic, refusals 99 to 3 out of 100 huggingface.co/llmfan46/gemma… ━━━━━━━━━━━━━━━━━━━━ 12gb · rtx 3060 12gb, 3080 12gb, 4070, 4070 super, 4070 ti, 5070 · amd rx 6700 xt, 6750 xt, 7700 xt · base bonsai 2 runs 50.1 tok/s on a 3060 12gb with my kernel and the speed head > 1. bonsai 2 27b abliterated v2 with the mtp speed head, 7.7gb, refusals 84% to 0% huggingface.co/BoldingBuilds/… > 2. qwen 3.8 27b abliterated squeezed to 10gb with gsq-rco, 2.12m downloads this month huggingface.co/huihui-ai/Huih… > 3. qwen 3.8 27b heretic gsq-rco, 9.3 to 10.1gb, refusals 0 to 1 out of 100 huggingface.co/0bserverx/Qwen… > 4. gemma 4 12b heretic, 7.4gb, 120k downloads huggingface.co/culturerevolt/… ━━━━━━━━━━━━━━━━━━━━ 16gb · rtx 4060 ti 16gb, 5060 ti 16gb, 4070 ti super, 4080, 5070 ti, 5080 · amd rx 6800, 6900 xt, 7600 xt, 7800 xt, 9060 xt 16gb, 9070 xt · base bonsai 2 runs 67.3 tok/s on a 5060 ti 16gb with everything on > 1. qwen 3.8 27b heretic ara, built for 16gb, 14gb huggingface.co/Bucoid/Qwen3.8… > 2. qwen 3.8 27b aggressive with the mtp head, 12.8 to 15.7gb, 0 refusals out of 465, 2.06m downloads this month huggingface.co/HauhauCS/Qwen3… > 3. bonsai 2 27b heretic with the full 262k window, the speed head and vision all on huggingface.co/OS-Software/Te… > 4. gemma 4 26b a4b abliterated, moe with 4b active per token, 13.4gb huggingface.co/groxaxo/Huihui… ━━━━━━━━━━━━━━━━━━━━ 24gb, the danger zone · rtx 3090, 3090 ti, 4090 · amd rx 7900 xtx, and the 20gb 7900 xt fits these too · base qwen 3.8 27b q4 runs 41.3 tok/s on my 3090 with the mtp head, owners report 61 on a 3090 ti and 76 on a 4090 > 1. qwen 3.8 27b uncensored, q4 16.8gb, 2.31m downloads this month huggingface.co/JonathanColett… > 2. qwen 3.8 27b abliterated, q4 16.8gb, 2.12m downloads huggingface.co/huihui-ai/Huih… > 3. qwen 3.8 27b aggressive with the mtp head, 0 refusals out of 465, 2.06m downloads huggingface.co/HauhauCS/Qwen3… > 4. qwen 3.8 27b heretic, refusals 0 to 1 out of 100 huggingface.co/0bserverx/Qwen… > 5. qwen 3.8 27b fable cold fusion, the uncensored coder merge, 2.09m downloads huggingface.co/DavidAU/Qwen3.… > 6. qwen 3.6 35b a3b aggressive, moe with 3b active, 11.9m downloads all time and the most liked on this list huggingface.co/HauhauCS/Qwen3… ━━━━━━━━━━━━━━━━━━━━ refusal numbers are each releaser's own test, so read them per model, and the speeds are the base models on nvidia cards, these builds keep the same size so they run the same. amd cards run the same files on llama.cpp's vulkan or rocm backends, the bonsai builds run on prismml's llama.cpp no refusals means no guardrails, you own what you ask. bookmark this, find your card and pull one tonight

好家伙！TradingView 里那些不公开在官方脚本库里的 Pine Script，有人专门整理成了一个开源仓库 PineScript，450+ Stars，MPL-2.0 开源，一套偏 ICT、动量交易和 Screener 的指标源码合集。 平时你在 TradingView 看到 FVG、Liquidity Sweep、Order Block、Market Structure 这些指标，很多只能看结果，很难知道背后到底怎么算。 这个仓库不一样，源码直接摆出来，复制进 Pine Editor 就能跑，还能继续拆逻辑、改参数、做自己的版本。 核心能力： 1️⃣ ICT 工具：FVG、Liquidity Sweep、MSS、Mitigation Block、Equal Highs / Lows 都有 2️⃣ 自带 Scanner：FVG、Order Block Retracement、Market Profile 可以直接拿来扫图 3️⃣ 动量策略：RSI Directional Momentum、Momentum Investing、Option Scalping 都有现成脚本 4️⃣ 趋势策略：包括 Sideways Market Skipper 等趋势跟随和震荡过滤思路 5️⃣ Market Profile：除了普通指标，还有 Market Profile + Screener 这种偏实战的工具 6️⃣ 源码全部可看：不是黑箱信号，计算逻辑和触发条件都能自己研究和修改 我觉得它最适合的用法，不是照搬某个信号去交易。 而是看到一个好指标以后，直接去看它的 Pine Script 到底怎么写，再拆成自己的交易逻辑。 对于 TradingView、Pine Script、ICT 或技术指标研究的人，这种源码库比单纯收藏指标更有价值。 GitHub： github.com/ArunKBhaskar/P…

爱吃折耳根的Ace @_zheergen

2 weeks ago

卧槽！股票、期货、期权、外汇、币圈，居然有人真把回测到实盘不用重写做成了一套专业交易引擎。 NautilusTrader，29K+ Stars，LGPL-3.0 开源，一套 Rust 核心 + Python 策略层的多资产量化交易系统。 很多量化项目真正麻烦的地方，是回测做完才发现：实时行情、订单执行、风控、Broker

4 14 64 42K 104

HOLY SH*T! The person who leaked this online is a real freaking madman! Now everyone can get 800 million tokens for AI agents for free The OmniRoute repository is blowing up GitHub. It is a local high-performance AI gateway that combines free limits from dozens of providers into a single continuous API stream Project architecture and key technical specifications: Interface and Integration: Provides a single OpenAI-compatible API endpoint. Natively connects to Cursor, VS Code (via Copilot / Cline / Aider) Claude Code, and custom scripts Connection Scale: Supports 350+ providers and 90+ services with free limits (Google AI Studio, Groq, Cerebras, SambaNova, OpenRouter, GitHub Models, Mistral AI, etc.) 18 Routing Strategies & Auto-Fallback: Custom switching system. On a ⁠429 Too Many Requests⁠ error or quota depletion, the gateway seamlessly redirects the request to the next model in the chain within milliseconds Dual Context Compression (RTK + Caveman): Built-in engines cut unnecessary logs, syntax, and message history by 15–95% saving token volume before sending the request Security (Local-first & 0 Telemetry): All API keys are locally encrypted using the AES-256-GCM algorithm. Data goes directly to providers, bypassing third-party clouds Agent Protocols (MCP & A2A): Full support for Model Context Protocol (MCP) and Agent-to-Agent architecture for automated model pool management Launch Options: CLI (⁠npm⁠), Docker container, Electron desktop app, and PWA github.com/diegosouzapw/O…

Bober_smart @Bober_smart

4 weeks ago

x.com/i/article/2103…

27 7 104 96K 121