Minara AI retweeted
Hermes Agent setup guide 👇
1. Get a key
minara.cloud/keys → Create API key (sk-...)
2. Add it to Hermes
echo 'MINARA_API_KEY=sk-your-key' >> ~/.hermes/.env
3. Register Minara as a provider in ~/.hermes/config.yaml
providers:
minara:
name: Minara
base_url: minara.cloud/v1
key_env: MINARA_API_KEY
default_model: anthropic/claude-opus-5.5
4. Launch
hermes --provider minara -m anthropic/claude-opus-5.5
(or run `hermes model` and pick Minara as your default)
5. Switch models mid-chat, same key
/model openai/gpt-6-sol
/model deepseek/deepseek-v4-pro
/model moonshotai/kimi-k3
Tip: model IDs need the vendor prefix (openai/..., anthropic/...).
Minara AI retweeted
You can now run autonomous agent workflows directly from this open source desktop app 🤯
It’s called Minara AI, an open-source CLI and MCP server built for agent tool calling and system execution.
> Run multi-step agent tasks natively in your shell
> Connect directly to Claude, Cursor, and any MCP client
> Built-in x402 protocol for machine-to-machine payments
> Self-correcting execution loop with command monitoring
Quick test:
Open the app and type "diagnose memory leak in production"
100% open source.
Minara AI retweeted
your uncle says gold, crypto twitter says bitcoin, finance bros say MSTR. i built a minara desktop dashboard to compare them and track what changes
first, what i'm actually comparing:
→ bitcoin: you own the asset itself
→ gold: usually you own it through a fund, not the metal
→ MSTR: you own shares of a company that holds bitcoin
that last one is where most people get confused. when you buy MSTR you're not just betting on bitcoin, you're also betting on the company, its debt and its decision to issue more shares
so i built a dashboard with all three over the same period: return, volatility and max drawdown, with the exact instrument and source labeled for every number
and the numbers show it. in the last run MSTR dropped 0.98% while bitcoin only dropped 0.33%, with a deeper drawdown too. the stock people buy "to get bitcoin exposure" fell about 3x harder than bitcoin itself. gold was the calmest of the three
then i set up a recurring check that runs on its own and tells me what changed since the last one, so i don't rebuild this every week
full walkthrough below 👇🏼
@minara
#MinaraAI #FinancialAI #AgenticTrading #AgenticFinance
🤝 Paid partnership
Minara AI retweeted
YOUR FRIEND TOLD YOU TO BUY NVDA. HERE'S HOW TO CHECK IF HE'S RIGHT BEFORE YOU PUT IN A SINGLE DOLLAR
We've all been there. Someone you trust says "just buy it", and you either go all in on vibes or spend hours reading earnings reports you don't fully understand.
So I tried something different. I asked Minara, an AI built only for investing, to do the homework a Wall Street analyst would do:
→ Pulled the latest earnings and filings, with the date of every source
→ Separated real facts from management guidance and assumptions
→ Gave me the bull case AND the bear case
→ Listed 3 things that would prove each side wrong
That last part is what changed how I think about it. Instead of asking "should I buy?", you start asking "what would make me change my mind?"
Then I opened its sources to check the numbers myself. Don't blindly trust your friend, and don't blindly trust an AI either.
Next, I opened Trading Co-pilot to review the entry, position size, and exit levels. On NVDA spot with $100: entry $229.54, target $234.50, invalidation $226.45.
Its call? Hold. Good company, bad entry today.
Full video below :)
#MinaraAI #FinancialAI #AgenticTrading #AgenticFinance
🤝 Paid partnership
Minara AI retweeted
Most swing traders check what a stock is doing and what onchain wallets are doing as two separate jobs
I built one playbook that checks both in the same window: a post-earnings pullback on AI semis, confirmed against whether smart-money wallets are already positioning in the correlated crypto plays.
Here is the full setup, idea to backtest to execution, in one chat.
Minara AI retweeted
Pretty shocked at how accurate the Minara agent has captured my trading styles when I imported my ChatGPT context into the desktop. This is pretty sick @minara
Minara AI retweeted
I’m glad to see a team like @minara take trading strategies seriously enough to build a real product around them.
Minara turns market ideas into coded strategies you can backtest and run. Strategy Studio is what stood out to me.
Pick an asset and timeframe, then combine indicators like RSI, Bollinger Bands, ROC, volume, MACD, or Donchian channels.
Add your leverage, stop loss, target, and backtest window. You can explain the setup in plain English, upload a chart, or paste code. Minara builds it, tests it, and can run it live.
If you don’t want to build from scratch, the Marketplace has strategies you can inspect and run.
The one that fits how I trade is Major Trend CTA 1.0, a 4H, 1x trend strategy across $BTC, $ETH, and $SOL.
At the time I checked, it showed a 642.06% backtested return, 26.74% max drawdown, a 1.55 Sharpe ratio, and 3.20% out-of-sample return.
The headline return is mostly backtested. The out-of-sample result is what matters now. No strategy works in every market, so start small and manage your risk.
Worth exploring if you’re into systematic trading, and you can get 500 free credits here: minara.ai/app/top-strategies…
Minara AI retweeted
Well actually this pic inspired Minara’s logo. 🥹
Didn't expect so many intentions on why Minara looks like Hermes... 😶🌫️
FYI Minara launched at Aug 15 2025 and this is our launch post. And before launch we've been using this profile photo for a longer time. :)
We understand the doubts and questions. Still, the fact that a less famous company applies similar design should not be a sin. 🙏
And all product feedback is welcomed. 🫡
Minara AI retweeted
Minara just went live with their ''Minara Harness''
their financial AI, they've built a very cool platform, and I'm deploying some capital into their strategies, honestly very cool concept
very proud to be supporting the Minara on growth with @LunarStrategy
time to accelerate ⚡️
Minara AI retweeted
Minara Harness @minara の機能紹介動画と、実機スクショ40枚の解説記事を作りました。
市場を調べる → 戦略を作る → 自動で動かす。
この流れがひとつのアプリでつながるのが面白い。
まずは動画で全体像を。
各画面の使い方や検証結果は、リプの記事にまとめています👇
Minara AI retweeted
Minara Harness, tested with three base models across six benchmarks:
- Leads on financial reporting with all three models.
- Tops web research and spreadsheet benchmarks with GPT-5.6 Luna Pro.
- Also evaluated on secure tool use, agent workflows, and skill-based tasks.
Minara AI retweeted
The 12 profitable Hyperliquid addresses behind our analysis, grouped by observed trading behavior:
High-turnover, two-sided execution
0x399965e15d4e61ec3529cc98b7f7ebb93b733336
0x7839e2f2c375dd2935193f2736167514efff9916
0x6ba889db7f923622d3548f621ecc2054b80c1817
0x03b9a189e2480d1e4c3007080b29f362282130fa
0xb07856ebcb6b37967eaf7c4b4e64dcce28c6de15
0xe4c6ae25959d7fc66cf2dd5965fb78c5e09c4048
0x523852be2db1a76a0e088ecbff32e849544054e5
0x04dbde5c9d1e6a3239c8ff8ee3fca386a4a3c605
Active intraday trading / scalping
0xc926ddba8b7617dbc65712f20cf8e1b58b8598d3
0x8c625ff57d8a4374784c7eff585dfdc42ccec974
0x77375a8c9d13bf79afb2a87f1b0ac1dfd5f5bf66
Concentrated directional trading
0x862dd8e68f30693e3d3c9daa42a440bc6d2a1f0c
Read more ↓
