Phoenix 5D๐ฅ retweeted
An estimated 100 million American homes are running on a router that nobody has logged into since the day it was plugged in.
Not once. Not to change a password, not to check for updates, not to look at a single setting.
And those same homes are paying $70, $80, even $100 a month for "fast" internet.
Then the video call freezes in the middle of a work meeting. Netflix buffers on a Friday night. The kids' games lag. The Wi-Fi "just feels slow."
So what does everyone do? They call the provider and yell. They upgrade the plan. They buy a $140 Wi-Fi extender. One household I know replaced their router twice in 18 months and spent more than $400 doing it.
None of it fixed the problem.
Not because the internet was too slow. Not because the new router was bad. Because every router ships on factory settings built for an "average home on an average day," crowded onto the same Wi-Fi channel as every neighbor, steering devices onto the wrong band, using the slowest default DNS, running firmware that is months or years out of date, and treating your work video call the same as someone's 40GB game download.
Then I sat down for 10 minutes in the admin panel. I didn't change the plan. I didn't buy new hardware. I didn't call the provider.
I changed 6 settings: channel, band steering, DNS, firmware, QoS, and one buried security toggle most people never touch.
Same router. Same plan. Same house. [Speeds nearly doubled / insert your real before and after speed test numbers].
"Your router was never too slow. It was never told how to manage your home. And nobody, not your provider, not the store, not the box it came in, was ever going to tell you the fix was free."
Here are the 6 settings I changed (and exactly how to do it) ๐งต
Phoenix 5D๐ฅ retweeted
๐จ BTC BREAKDOWN: The $65K Trap & The Liquidity Sweep
1๏ธโฃ ๐ฏ Executive Conviction & Strategic Bias
Bitcoin is currently caught in a violent "liquidity vacuum." Price has broken below critical structural support, triggering a cascade of long liquidations. However, the tape is flashing a massive Bullish Spring (SMC Liquidity Sweep) with 86% confidence. We are seeing classic institutional accumulation behavior: price is being pushed down to flush out weak retail longs before a violent re-claim of the range. The bias is Neutral-to-Bullish on the dip, but strictly waiting for a reclaim of the $66,000 zone to confirm the sweep.
2๏ธโฃ ๐ข Fundamental Health & Valuation Matrix
* Market Cap $1.31 Trillion
* Asset Class: Decentralized Digital Gold / Smart Contract Platform
* Supply Dynamics: Fixed 21M cap. Post-halving supply shock is currently absorbing institutional ETF inflows.
* Macro Catalyst: Spot ETFs (IBIT, etc.) continue to act as the primary bid, but corporate treasury demand (MSTR) is currently paused due to price weakness.
3๏ธโฃ ๐ฌ Market Physics & Technical Structure
* Spot Price: $64,852.85 (-2.75% today)
* 20-Day MA: $75,825.57 (Overhead Resistance โ +16.9% above spot)
* 50-Day MA: $80,151.06 (Major Overhead Resistance โ +23.6% above spot)
* RSI (14): 21.6 (Deeply Oversold โ Extreme selling pressure)
* **RVOL**: 0.94x (Volume is actually *below* average, indicating a lack of aggressive new selling, just passive liquidation flow)
* ATR (14-day): $2,821.25 (Expected daily swing range is ~$2.8k)
* Bollinger Latent Energy: 35.0% Compression (Normal volatility, coiling for expansion)
* Key Support: $61,946.17 (Calculated Technical Support)
* Key Resistance: $81,955.00 (Calculated Technical Resistance)
4๏ธโฃ ๐ Smart Money Flow & Institutional Tape
* Dealer Gamma Exposure (GEX): **Zero-GEX Flip @ $39.84** (Note: This is relative to the $40 IBIT ETF price, not BTC spot. For BTC, dealer hedging is currently amplifying the downside as price falls).*
* Coinbase Premium Index: -0.0571% (MILD_BEARISH). US spot demand is slightly lagging global perps.
* Hyperliquid Perps OI: $6.92B. High open interest means the liquidation cascade is still active. If OI drops rapidly while price stabilizes, the flush is complete.
Phoenix 5D๐ฅ retweeted
You can basically build a tiny AI podcast studio on your laptop now with just this stack:
1. Whisper
github.com/openai/whisper
2. Demucs
github.com/facebookresearch/โฆ
3. Chatterbox
github.com/resemble-ai/chattโฆ
4. AudioCraft
github.com/facebookresearch/โฆ
5. Auto-Editor
github.com/WyattBlue/auto-edโฆ
Save this and set it up when you have time
Best Uncensored AI models now :
Overall : Qwen3.8-27B Abliterated
Small GPU : Gemma 4 12B Abliterated
Fastest : Gemma 4 26B-A4B Abliterated
Agentic : Qwen3.6-35B-A3B Uncensored
Writing : Mistral Small 3.2 Abliterated
No surgery : Hermes 4.3-36B
Image : FLUX.1-dev
Video : Wan 2.2
Ceiling : GLM-5.1 Abliterated (754B, not for home)
Where to get them :
all LLMs โ huggingface.co/huihui-ai (search the name, grab the GGUF)
Hermes โ huggingface.co/NousResearch
FLUX / Wan โ ComfyUI + the model's HF page
Setup :
ollama pull hf.co/<repo>:Q4_K_M
ollama run <model>
/set nothink
that's it.
no API key, no account, no card, just few clicks setup
more on how to setup below:
Here's 4 steps to run an uncensored AI on your own GPU for $0:
I run a local abliterated Gemma 4 for anything I don't want leaving my machine
By the end, you'll know how to:
- run a model which has no refuses
- plug it into Claude Code
- do it all for $0, no API key, no account
So, let's go step by step:
1. Install Ollama
curl -fsSL ollama.com/install.sh | sh
you want a recent version, older builds don't speak the Anthropic API and don't support Gemma 4
2. Pull the model
ollama pull igorls/gemma-4-12B-it-heretic-GGUF
the biggest mistake people make here is grabbing a random "uncensored" tag with no quality numbers
this one is abliterated with Heretic. refusals went 99/100 โ 0/100 at a KL divergence of 0.028 from the original, so you keep almost all of the base model's capability
one thing: it's still Google's Gemma license underneath, fine for personal use, but read the terms before you build a product on it
3. Turn off thinking (the part every other guide skips)
Gemma 4 is a hybrid thinking model and the abliteration targets non-thinking mode
leave thinking on and the reasoning block eats your token budget and the answer gets cut off
ollama run igorls/gemma-4-12B-it-heretic-GGUF
/set nothink
if you're calling the API instead, add "think": false to the request
4. Plug it into Claude Code (OPTIONAL)
ollama launch claude --model igorls/gemma-4-12B-it-heretic-GGUF
or manually:
export ANTHROPIC_BASE_URL=http://localhost:11434
export ANTHROPIC_AUTH_TOKEN=ollama
claude --model igorls/gemma-4-12B-it-heretic-GGUF
two gotchas:
- bump the context first. Ollama defaults to 4k and Claude Code's system prompt alone is bigger than that, so the model never sees its tool instructions. restart the server with OLLAMA_CONTEXT_LENGTH=32768 ollama serve (setting it in your shell does nothing if Ollama is already running as a service). this needs more VRAM than the 8GB below
- Gemma 4 tool calling is still rough, you should expect some failed edits, imo for serious agentic work use qwen3.5 or glm-4.7-flash instead
What you need:
- 8-12GB VRAM for Q4 (12B)
- ~8GB of disk for Q4, ~13GB for Q8
- more of both if you want a big context
honest take:
- "abliterated" means refusals removed, not quality added. it's still a 12B
- fewer refusals โ smarter. for real coding work I still use cloud models
- you own the weights, your prompts never leave the box. that's the actual win
*IMPORTANT: don't use it for anything illegal or harmful. that's on you, not the model
now i'm collecting a list of every uncensored model i'm adding to my local stack and which tasks each one replaces for me
also planning a big "Uncensored 101" article, still figuring out all the angles
save it so you don't lose it
Phoenix 5D๐ฅ retweeted
NVIDIA Open-Sourced Auto Ontology, a Semantic Layer for AI Agents!
Auto Ontology is a Python framework for building semantic, graph-based reasoning layers over data.
AI agents can already query databases and retrieve documents.
The harder problem is making sure they understand what the data actually means.
A metric like โrevenueโ might be defined differently across two systems. The same customer can appear under different identifiers. A generated SQL query can be syntactically correct and still use the wrong business definition.
Auto Ontology is built for that layer in between.
It can build a semantic catalog over enterprise metadata so agents can reason about entities, metrics, relationships, lineage, and where trusted definitions come from.
The workflow is roughly:
Enterprise data + metadata
โ Auto Ontology
โ Semantic catalog / knowledge graph
โ AI agent
โ Grounded query + provenance
It can generate ontologies from existing metadata, connect to semantic models you already have, reconcile inconsistent definitions across systems, and expose that shared context through MCP, CLI, or API.
For structured data, it can also translate natural-language questions into SQL while validating the query against the ontology and certified definitions before returning the result.
So instead of an agent simply asking:
โHow much revenue did we generate last month?โ
it can understand which revenue definition to use, which tables contain the right data, how those entities relate, and where the final answer came from.
Auto Ontology currently supports relational sources including Databricks, Postgres, Snowflake, and DuckDB.
It also exposes an MCP server, which means agents can query the same semantic layer directly.
For enterprise agents, retrieval alone is not enough.
They also need a shared understanding of what the underlying data actually means.
100% open source.
Iโve shared the GitHub repo in the comments!
NVIDIA Open-Sources Auto Ontology, a Semantic Layer for AI Agents
Your boss asks the AI agent: "What was revenue last quarter?"
It answers in 3 seconds. Clean number. Nice chart.
But it used the wrong "revenue" column. ๐ก๐ผ๐ฏ๐ผ๐ฑ๐ ๐ป๐ผ๐๐ถ๐ฐ๐ฒ๐.
That's the problem NVIDIA's new open-source library, Auto Ontology, goes after. Its README puts it plainly:
"โฆ๐ข ๐ณ๐ฆ๐ท๐ฆ๐ฏ๐ถ๐ฆ ๐ฎ๐ฆ๐ต๐ณ๐ช๐ค ๐ข๐ฏ๐ด๐ธ๐ฆ๐ณ๐ฆ๐ฅ ๐ง๐ข๐ด๐ต๐ฆ๐ณ ๐ช๐ด ๐ธ๐ฐ๐ณ๐ต๐ฉ ๐ฏ๐ฐ๐ต๐ฉ๐ช๐ฏ๐จ ๐ช๐ง ๐ต๐ฉ๐ฆ ๐ฏ๐ถ๐ฎ๐ฃ๐ฆ๐ณ ๐ช๐ด ๐ธ๐ณ๐ฐ๐ฏ๐จ."
๐ช๐ต๐ฎ๐ ๐ถ๐ ๐ถ๐:
A "context layer" that sits between an AI agent and your database. It tells the agent where the data lives, what each field means, and whether to use it.
๐๐ผ๐ ๐ถ๐ ๐๐ผ๐ฟ๐ธ๐, ๐๐๐ฒ๐ฝ ๐ฏ๐ ๐๐๐ฒ๐ฝ:
โข ๐๐ถ๐๐ฐ๐ผ๐๐ฒ๐ฟ: it reads your database metadata and drafts an ontology, a map of business terms like "active customer"
โข ๐๐ฟ๐ผ๐๐ป๐ฑ: agents use those terms instead of guessing from raw table names
โข ๐ฉ๐ฎ๐น๐ถ๐ฑ๐ฎ๐๐ฒ: it checks generated SQL against certified definitions
โข ๐๐
๐ฝ๐ผ๐๐ฒ: agents reach it through MCP, a CLI or an API
โข ๐๐ฒ๐ฎ๐ฟ๐ป: certified corrections go back into the map
๐๐ผ๐ฟ ๐๐๐ฒ๐ฟ๐:
Ask a question, get ๐๐ต๐ฒ ๐ฎ๐ป๐๐๐ฒ๐ฟ, ๐๐ต๐ฒ ๐ฆ๐ค๐ ๐ถ๐ ๐ฟ๐ฎ๐ป, ๐ฎ๐ป๐ฑ ๐๐ต๐ฒ ๐ฟ๐ผ๐๐. You can check the work, not just trust it.
๐๐ผ๐ฟ ๐ฏ๐๐ถ๐น๐ฑ๐ฒ๐ฟ๐:
โข Apache 2.0, runs with Docker Compose or Helm
โข Works with Databricks, Postgres, Snowflake and DuckDB
โข Defaults to NVIDIA's Nemotron models, but ๐ฎ๐ป๐ ๐ข๐ฝ๐ฒ๐ป๐๐-๐ฐ๐ผ๐บ๐ฝ๐ฎ๐๐ถ๐ฏ๐น๐ฒ ๐ฒ๐ป๐ฑ๐ฝ๐ผ๐ถ๐ป๐ ๐๐ผ๐ฟ๐ธ๐
โข The MCP server has ๐ต ๐ฟ๐ฒ๐ฎ๐ฑ-๐ผ๐ป๐น๐ ๐๐ผ๐ผ๐น๐. There is deliberately no tool to browse raw tables, so agents must go through the business terms
๐๐ ๐ฏ๐๐ถ๐น๐ฑ๐ ๐ผ๐ป ๐ฒ๐ฎ๐ฟ๐น๐ถ๐ฒ๐ฟ ๐๐ผ๐ฟ๐ธ:
โข Cube and dbt MetricFlow: semantic layers you write by hand in code
โข Wren AI: an open semantic layer for AI agents, with a context layer built partly by agents
โข Snowflake's Semantic View Autopilot: also drafts semantic models with AI, but only for Snowflake
What's different here: ๐ผ๐ฝ๐ฒ๐ป ๐๐ผ๐๐ฟ๐ฐ๐ฒ, ๐๐ผ๐ฟ๐ธ๐ ๐ฎ๐ฐ๐ฟ๐ผ๐๐ ๐ณ๐ผ๐๐ฟ ๐ฑ๐ฎ๐๐ฎ๐ฏ๐ฎ๐๐ฒ๐, ๐ฎ๐ป๐ฑ ๐ฏ๐๐ถ๐น๐ ๐ณ๐ผ๐ฟ ๐ฎ๐ด๐ฒ๐ป๐๐ ๐๐ต๐ฟ๐ผ๐๐ด๐ต ๐ ๐๐ฃ.
๐ง๐ต๐ฒ ๐ต๐ผ๐ป๐ฒ๐๐ ๐ฝ๐ฎ๐ฟ๐:
โข ๐ง๐ต๐ฒ "๐ต๐ฌ% ๐ณ๐ฎ๐๐๐ฒ๐ฟ" ๐ฐ๐น๐ฎ๐ถ๐บ ๐ถ๐ ๐ก๐ฉ๐๐๐๐'๐ ๐ผ๐๐ป ๐๐ฒ๐๐๐ถ๐ป๐ด, with no baseline published
โข ๐ก๐ผ ๐ฎ๐ฐ๐ฐ๐๐ฟ๐ฎ๐ฐ๐ ๐๐ฐ๐ผ๐ฟ๐ฒ๐ ๐ฎ๐ฟ๐ฒ ๐ฝ๐๐ฏ๐น๐ถ๐๐ต๐ฒ๐ฑ. NVIDIA ships an eval harness (BIRD, 1,534 questions) but no results
โข That harness was tested with ๐๐น๐ฎ๐๐ฑ๐ฒ ๐ข๐ฝ๐๐ ๐ฐ.๐ด, ๐ป๐ผ๐ ๐ก๐ฒ๐บ๐ผ๐๐ฟ๐ผ๐ป
โข ๐ช๐ถ๐๐ต๐ผ๐๐ ๐๐ฎ๐๐ต๐ถ๐๐ผ๐ฟ๐ฝ ๐ฉ๐ฎ๐๐น๐, ๐ฑ๐ฎ๐๐ฎ๐ฏ๐ฎ๐๐ฒ ๐ฝ๐ฎ๐๐๐๐ผ๐ฟ๐ฑ๐ ๐ฎ๐ฟ๐ฒ ๐๐๐ผ๐ฟ๐ฒ๐ฑ ๐๐ป๐ฒ๐ป๐ฐ๐ฟ๐๐ฝ๐๐ฒ๐ฑ in Postgres (per the README)
โข It's young: ๐ฎ๐ฏ๐ผ๐๐ ๐ณ๐ฑ๐ฌ ๐๐ถ๐๐๐๐ฏ ๐๐๐ฎ๐ฟ๐, and NVIDIA isn't accepting outside contributions
๐ง๐ต๐ฒ ๐ฏ๐ถ๐ด๐ด๐ฒ๐ฟ ๐ฝ๐ผ๐ถ๐ป๐: the hard part of AI analytics isn't writing SQL. ๐๐'๐ ๐ฎ๐ด๐ฟ๐ฒ๐ฒ๐ถ๐ป๐ด ๐ผ๐ป ๐๐ต๐ฎ๐ ๐๐ต๐ฒ ๐๐ผ๐ฟ๐ฑ๐ ๐บ๐ฒ๐ฎ๐ป.
GitHub: github.com/NVIDIA/auto-ontolโฆ
Links in reply ๐
Phoenix 5D๐ฅ retweeted
holy f*ck, NO HUMAN WEB DESIGNER WANTS TO COMPETE WITH THIS IN 2027.
youโre watching AI agents build websites at a speed most humans canโt match.
and the scary part? these tools are available to almost anyone.
7 GitHub repos you NEED to learn before your clients find someone who already has:
1. Bolt.diy - BUILD ENTIRE WEBSITES WITH AI
generate, edit and run full-stack web apps with AI.
github.com/stackblitz-labs/bโฆ
2. shadcn/ui - PROFESSIONAL UI IN MINUTES
beautiful, customizable components for modern websites.
github.com/shadcn-ui/ui
3. Magic MCP - YOUR AI FRONTEND DESIGNER
generate polished React components directly from your AI coding workflow.
github.com/21st-dev/magic-mcโฆ
4. Playwright MCP - AI THAT TESTS ITS OWN WORK
let AI agents open websites, click buttons and catch UI bugs.
github.com/microsoft/playwriโฆ
5. Chrome DevTools MCP - FIX BROKEN WEBSITES FASTER
give AI access to browser debugging, performance and network diagnostics.
github.com/ChromeDevTools/chโฆ
6. Vercel Agent Skills - MAKE AI CODE PRODUCTION-READY
React optimization, frontend best practices and web design reviews.
github.com/vercel-labs/agentโฆ
7. Webstudio - BUILD VISUALLY, KEEP CONTROL
an open-source visual website builder with reusable components and styling.
github.com/webstudio-is/websโฆ
the real threat isnโt AI replacing every designer.
itโs ONE person using AI to deliver what used to take an entire team.
2027 is getting closer.
learn these tools before your competitors do.
LIKE + BOOKMARK + REPOST this so your designer friends see it.
and be honest in the comments:
do you know anyone whoโs already LOST THEIR JOB because of AI?
Phoenix 5D๐ฅ retweeted
holy f*ck, I just sold this single AI-generated 3D neuron for $2,000 to a private biotech lab.
a private lab saw my visuals on X and reached out. they just ordered a prototype for their modeling work.
designers who can't do this will be looking for a job in 2027. save this before its too late:
7 design repos that make Claude Code feel like a whole design team:
1. Awesome DESIGN.md
84 design systems pulled from real sites, each one a single markdown file. drop it in your project, say "build me a page that looks like this" and it actually does
github.com/VoltAgent/awesomeโฆ
2. Hallmark
a design skill from Together AI that refuses to look AI-generated. 50 themes, builds the page around what the visitor came to do, then checks it in a real browser. /hallmark redesign rebuilds any old page
github.com/Nutlope/hallmark
3. Huashu Design
type one sentence, get a launch animation, a clickable app prototype, an editable PPT or an infographic. it shows you 6 directions first, you pick 3, it builds them for real
github.com/alchaincyf/huashuโฆ
4. video-shotcraft
tell Claude "make a promo for my product". it screenshots your real pages, storyboards, animates, syncs to the beat, adds sound and checks every frame. you get an MP4 you can keep editing
github.com/Vincentwei1021/viโฆ
5. OpenPencil
an open-source Figma alternative that opens real .fig files. describe what you want in chat and the AI builds the design, and Claude Code can drive it directly
github.com/open-pencil/open-โฆ
6. Motion2Code
show it a motion design video and it rebuilds it as clean SVG + GSAP code, checked frame by frame against the original. a 10-second animation ends up around 24 KB
github.com/lovisdotio/motionโฆ
7. Blender Agent Skills
22 pro Blender workflows for Claude Code: modeling, materials, rigging, animation, lighting, rendering. the full 3D pipeline, not just "make a cube"
github.com/EL4CTEO/blender-sโฆ
the neuron in this video?
I built it myself with Claude Code and a few 3D and motion tools and sold it for 2K. no studio, no team.
want to make something like this?
like + save this post, comment "NEURON" or DM me and I'll send you the exact prompt.
Phoenix 5D๐ฅ retweeted
The cloud hyperscalers just delivered their verdict on AI infrastructure demand: $109.9 Billion in combined quarterly revenue.
โIn freshly reported Q3 numbers, cloud growth re-accelerated across all three giants:
โข AWS: Grew 39.3% to $46.0B ($184B annualized run-rate) with $28.1B in operating income (Amazon IR)
โข Microsoft Azure: Surged 48% at constant currency to $33.2B in its first official dollar disclosure (Microsoft)
โข Google Cloud: Rose 103% to $30.7B, tripling operating profit to $11.1B
โThe take-or-pay enterprise contracts are converting into recognized sales. The big question for Wall Street now shifts from revenue proof to return on capital expenditure.
โWhich cloud platform is capturing the highest-margin enterprise agent workloads?
โ#CloudComputing #AWS #Azure #GoogleCloud #TechEarnings #AI
๐ค Made with AI
Phoenix 5D๐ฅ retweeted
Core PCE came in hot at 3.12% y/y with Q3 GDP beating forecasts โ one day after the Fed's 10โ2 hold. Traders now price 81% odds of a Dec 9 hike. The inflation fight is officially back on the calendar. (investors.com; CME FedWatch via longbridge)
Phoenix 5D๐ฅ retweeted
5 GITHUB REPOS FOR GIVING AI EYES + HANDS ON YOUR COMPUTER:
1.โ โ OmniParser
github.com/microsoft/OmniParโฆ
2.โ โ Agent-S
github.com/simular-ai/Agent-โฆ
3.โ โ UI-TARS Desktop
github.com/bytedance/UI-TARSโฆ
4.โ โ OpenAdapt
github.com/OpenAdaptAI/OpenAโฆ
5.โ โ screen-use
github.com/tongriyaotxt/screโฆ
AI gets a lot more interesting when it can actually see your screen and operate the apps on it
Phoenix 5D๐ฅ retweeted
Tech giants are spending $1 Trillion on AI infrastructure. Here is the exact equation that determines whether the market booms or corrects.
โTrillions in capital expenditure are flowing into power grids, liquid-cooled data centers, and GPU clusters (Celent). For valuations to hold, the left side of the equation must unlock the right:
โโข The Inputs ($1T Capex): High-voltage grid upgrades, hyperscale compute clusters, and sovereign data centers
โข The Outputs ($1.5T+ Value Capture): Enterprise margin expansion, autonomous back-office cost reduction, and a 2.5x surge in workforce productivity leverage
โWatch utility bottlenecks and enterprise software net-retention ratesโthey are the canaries in the coal mine for the entire macro market.
โ#MacroEconomics #TechStocks #CapitalMarkets #DataCenters #InvestingStrategy
๐ค Made with AI
Phoenix 5D๐ฅ retweeted
Neocloud stocks retreat after Musk's 'guppy-level death' warning
Nebius, CoreWeave and Iren each dropped more than 4% before Monday's market open. Musk said debt markets have almost entirely shut off financing for companies without revenue beyond AI data centers.
tradingview.com/news/seekingโฆ
Phoenix 5D๐ฅ retweeted
the fastest growing GitHub repos in finance + AI this week:
1. Kronos (+2,653 โ
)
open source foundation model for candlestick charts, trained on data from 45+ exchanges. you feed it price history, it forecasts the next bars. code hasn't changed since april, people just found it.
github.com/shiyu-coder/Kronoโฆ
2. TradingAgents (+1,944 โ
)
multi-agent LLM trading framework: analysts, researchers, a trader and a risk team arguing before every trade. #2 again, now 113K stars.
github.com/TauricResearch/Trโฆ
3. day-trader-skill (+972 โ
)
11 Claude skills for day trading. chart scanner, news tracker, entry and stop planner, and a "time machine" that replays the session 10,000 times before you trade. paper trading by default. 4 days old.
github.com/Jakeschincariol/dโฆ
4. Vibe-Trading (+821 โ
)
personal trading agent from HKUDS. research in plain english, strategy drafts, backtests, MCP support.
github.com/HKUDS/Vibe-Tradinโฆ
5. last30days-skill (+730 โ
)
agent skill that researches any topic across Reddit, X, YouTube, HN and Polymarket, and ranks results by upvotes, likes and real money.
github.com/mvanhorn/last30daโฆ
6. daily_stock_analysis (+487 โ
)
LLM stock analysis for US, China, HK and more. sends you a daily decision dashboard to Telegram, Discord or email.
github.com/ZhuLinsen/daily_sโฆ
7. FinceptTerminal (+412 โ
)
open source Bloomberg-style terminal. market data, research and AI agents in one desktop app.
github.com/Fincept-Corporatiโฆ
8. freqtrade (+360 โ
)
the OG open source crypto trading bot, with a machine learning module (FreqAI). backtests, Telegram control, supports Hyperliquid.
github.com/freqtrade/freqtraโฆ
9. tick-stock-panel (+346 โ
)
self-hosted quant workbench for China A-shares. LLM builds your strategy, then backtest and monitoring. 65 API endpoints + an MCP server.
github.com/shy3130/tick-stocโฆ
10. OpenBB (+294 โ
)
open data platform for analysts, quants and AI agents. stocks, options, crypto, macro. has an MCP server.
github.com/openbq-org/OpenBB
bookmark this and start today.
the fastest growing GitHub repos in finance + AI this week:
1. aurelio-finance (+2,388 โ
)
personal finance app with an AI advisor you can talk to. tracks net worth, ETFs, investments and cash. runs on your own computer, AI is optional with your own key. the repo is 11 days old.
github.com/LosaLosSantos/aurโฆ
2. TradingAgents (+1,244 โ
)
multi-agent LLM trading framework, analysts + researchers + trader + risk team in one stack. back on the list again. now 111K stars, new release lets it run on your ChatGPT Plus/Pro plan.
github.com/TauricResearch/Trโฆ
3. last30days-skill (+735 โ
)
agent skill that researches any topic across Reddit, X, YouTube, HN and Polymarket, and ranks what it finds by upvotes, likes and real money. works in Claude Code, Codex and Cursor.
github.com/mvanhorn/last30daโฆ
4. Vibe-Trading (+621 โ
)
personal trading agent from HKUDS. market research in plain english, strategy drafts, backtests, MCP support. one command to give your agent trading skills.
github.com/HKUDS/Vibe-Tradinโฆ
5. nautilus_trader (+463 โ
)
Rust trading engine where the same strategy code runs in backtest and live. any asset, any venue with an API. biggest day of the week was friday, +130.
github.com/nautechsystems/naโฆ
6. daily_stock_analysis (+428 โ
)
LLM stock analysis for US, China, HK, Japan, Korea and Taiwan. sends you a daily decision dashboard to Telegram, Discord, Slack or email.
github.com/ZhuLinsen/daily_sโฆ
7. freqtrade (+342 โ
)
the OG open source crypto trading bot, now with a machine learning module (FreqAI). backtesting, Telegram control, supports Hyperliquid.
github.com/freqtrade/freqtraโฆ
8. OpenBB (+317 โ
)
open data platform for analysts, quants and AI agents. stocks, options, crypto, macro. one pip install, has an MCP server.
github.com/openbq-org/OpenBB
9. FinceptTerminal (+293 โ
)
open source Bloomberg-style terminal. market analytics, research and AI agents in one desktop app.
github.com/Fincept-Corporatiโฆ
10. QuantDinger (+266 โ
)
self-hosted AI trading OS. idea โ python strategy โ backtest โ paper โ live, with monitoring. was #6 back in may, still growing.
github.com/OpenByteInc/Quantโฆ
bookmark this and start today.
Phoenix 5D๐ฅ retweeted
It is becoming f*cking dangerous: a model that never writes a single word now judges every move your Claude AI agent makes
-> Create a folder named JEV LAB on your PC and launch Claude Code inside it
-> Install the Jev CLI: npm install -g jev-axi
-> Grab a key at console.typesafe.ai and run: jev-axi config set apiKey YOUR_KEY
-> Run jev-axi setup safety so every command Claude runs gets checked first. Dangerous downloads get denied, routine tests pass for free
-> Run jev-axi setup git-hooks to stop any commit that sneaks in a password or API key
-> Feed it anything your agent reads: jev-axi guard < email.txt
Hidden "ignore your previous instructions" tricks get flagged in under half a second
-> Run jev-axi setup supervise --project to catch the moment your agent quits early or gets stuck in a loop
The whole setup takes around 5 minutes
Turn Claude into an agent with a second opinion: every command screened, every page checked for prompt injection, every commit scanned for secrets. For a fraction of a cent per check
Bonus: pipe a broken build into jev-axi triage and it finds the root cause for you
Bookmark this article and check repo below:
github.com/shiftynick/jev-axโฆ
Phoenix 5D๐ฅ retweeted
๐จ ESTOS 10 REPOS DE TRADING ESTรN CREANDO MILLONARIOS EN SILENCIOโฆ Y SON 100% GRATIS
> 10 repos open source
> 435K+ estrellas en GitHub
> $0 en licencias
> 1 stack: pensar, probar, ejecutar
10 repos de trading open source que deberรญas tener guardados:
AGENTES DE IA QUE ANALIZAN:
1. TradingAgents (111K โญ) โ github.com/TauricResearch/Trโฆ
un equipo de agentes de IA: analistas, investigadores alcistas y bajistas, trader y gestor de riesgo que debaten cada operaciรณn
2. ai-hedge-fund (64K โญ) โ github.com/virattt/ai-hedge-โฆ
agentes que analizan acciones como lo harรญan Buffett, Munger o Burry. Educativo: no opera con dinero real
BOTS QUE OPERAN:
3. freqtrade (55K โญ) โ github.com/freqtrade/freqtraโฆ
el bot de cripto gratis mรกs usado: backtesting, optimizaciรณn de estrategias y control desde Telegram
4. Hummingbot (20K โญ) โ github.com/hummingbot/humminโฆ
bots de market making y arbitraje en exchanges de cripto
MOTORES DE BACKTESTING:
5. NautilusTrader (30K โญ) โ github.com/nautechsystems/naโฆ
motor escrito en Rust: la misma estrategia corre en backtest y en vivo
6. Lean de QuantConnect (22K โญ) โ github.com/QuantConnect/Lean
el motor de trading algorรญtmico de QuantConnect, en Python y C#
IA PARA QUANTS:
7. Qlib de Microsoft (49K โญ) โ github.com/microsoft/qlib
plataforma de inversiรณn cuantitativa con IA, de los datos al modelo
8. FinRL (16K โญ) โ github.com/AI4Finance-Foundaโฆ
aprendizaje por refuerzo aplicado a trading
LA BASE:
9. ccxt (44K โญ) โ github.com/ccxt/ccxt
una sola API para mรกs de 100 exchanges de cripto y mercados de predicciรณn
10. machine-learning-for-trading (21K โญ) โ github.com/stefan-jansen/macโฆ
todo el cรณdigo del libro de Stefan Jansen: de conseguir datos a operar en vivo
Guรกrdalo antes de que sea tarde ๐
FULL TUTORIAL TO BUILD YOUR OWN AI TRADING BOT WITH CLAUDE IN 14 MINUTES
This ex-Meta engineer builds a bot that trades for you 24/7, and you donโt need to know how to code.
โ connects Claude Code to Alpaca with an MCP
โ gives it a simple strategy (20/50-day moving average crossover)
โ the bot checks the market on its own and buys or sells when the signal hits
people sell courses like this for $497. heโs giving it away for free, and he tests everything with paper trading first.
๐ Save it, itโs worth it.
ex-Citadel Quant just revealed his Opus 5.5 & Fable 5.5 trading system
this shouldn't be f*cking public...
It turns Claude into a 24/7 research & trading swarm of AI agents: one researches, one trades the edge, one can kill any position
Send it to Opus 5.5 , and enjoy. Then read how to build Bloomberg terminal for ~300$:
Three prompts turn Grok + Opus 5.5 into a Citadel-level research desk that works for you 24/7
Ken Griffin spent 36 years and billions building this machine. you build a copy in one evening, for free
one agent hunts for ideas. one makes the calls. one tears every idea apart before real money touches it
and none of them holds the keys to the vault. risk limits are hard-coded, no model can override them. that's what every blown-up bot was missing
no code, no quant background. drop three ready prompts into three agents, set your risk limits, done
all three are in the paper below. worth saving
Ken Griffin, who runs Citadel, just admitted that the most expensive skill of his quants, the one he paid them millions for, now costs nothing
And for the first time in history, that skill belongs not just to a hedge fund, but to anyone with access to AI
A fund's entire quant team runs one cycle: come up with a strategy, turn it into code, test it on history, deploy it to the market
Every step used to cost separate people and years of training. Now AI does the whole cycle
You describe an idea in words, it writes the strategy itself, runs it on years of real data, and if it's alive, deploys it straight to a broker
But the speed isn't the point. The point is what the retail trader doesn't understand
You take a strategy with a beautiful chart and you believe it. But a beautiful chart is no proof
Run enough versions and a profitable curve will appear even on random numbers. One just got lucky, and you decided you found gold
Funds don't trust any curve until they test it on data the strategy has never seen. There the fit falls apart, and a real edge survives
I ran two of the most-cited quant papers through this cycle
Time Series Momentum on Bitcoin: +1,537%. Alive
Betting Against Beta, a classic from every textbook: minus 7.78% a year while the market rose 14.86%. A corpse
One still prints money, the other died long ago. And trading blind, you'd never know which is which until it took your deposit
If you want to run your own idea through the same test, it takes minutes here (free trial after signing up): join.horizon.trade/qwinsiiii
The entire hedge fund is six steps. AI already ate four of them. Which ones, I broke down in the article below
Phoenix 5D๐ฅ retweeted
ANTHROPIC JUST DROPPED THE BEST 1-HOUR AI AGENT CRASH COURSE THAT YOUโLL FIND
It covers:
- building /skills, /loops, and graphs
- โ agent dreaming & orchestration
- reverse engineering with Claude agents
Phoenix 5D๐ฅ retweeted
20+ GitHub repositories for algorithmic trading with Python.
Grab the repo here:
Phoenix 5D๐ฅ retweeted
Ex-Jane Street intern openly shared in private quants forum, how he used free REA MCP to reverse-engineer exchange fee structures and built a solo market-making bot doing $27k/month.
feed this exact prompt into Claude code and you'll have a working evidence-backed trading engine.
Bookmark it right now. 53k upvotes




































