La plus long éclipse totale de votre vie aura lieu l'année prochaine, le 2 août 2027, avec 6min23 de totalité.
Ni nous, ni probablement nos enfants n'en verront plus une aussi longue.
Il faudra attendre l'an 2114 pour en avoir une plus longue.
A senior Anthropic engineer just dropped 11-page PDF on "Loop Engineering" for agentic systems.
The shift: you stop prompting the agent. You build the system that prompts it instead.
Schedule → Discover → Build → Verify → Repeat
Every loop runs one turn, five moves:
• Discovery: it finds its own work - failing CI, open issues, recent commits - instead of being handed a list.
• Handoff: each task gets an isolated git worktree so parallel agents don't collide.
• Verification: a second agent, told to assume the code is broken, reviews the first. The "thing that can say no."
• Persistence: results get written to disk, never left in a context window that gets flushed.
• Scheduling: an automation wakes it on a timer. That's what makes it a loop.
The key insight: an agent grading its own work always praises it.
This 11-page PDF changed how I'm building agentic systems today.
Read it now, then explore the article below.
Millions of books and papers, freely accessible.
Anna's Archive is the world's largest shadow library - invaluable for investigators who hit paywalls. 📚
Free to use. Learn more: https://t.co/GVrpXnp8tt
how to claim $81 worth gemini pro for free
you can use gemini pro free for 4 months
> go to: https://t.co/r1Eg9A9SOM
- connect a gmail where you didn't have gemini pro previously
- connect your card or you can add any of the payment options available
> dont forget to remove the payment options, just before the offer ends after 4 months
> if you fail to remove, then they will take money from you
> few benefits of gemini pro
- it has a 2,000,000 token context window (big memory)
- good for coding and good reasoning ability
- it can surf the web realtime and come up with good results
the fastest growing GitHub repos in finance this week:
1. TradingAgents (+3,822 ★)
multi-agent LLM trading framework built for financial research and execution. combines analyst agents, sentiment models, portfolio reasoning, and provider integrations into a single trading stack.
2. AI-Trader (+2,434 ★)
fully automated agent-native trading system. built around autonomous decision-making, price fetching, execution, and monitoring workflows. focused on end-to-end AI-driven trading infrastructure.
3. scientific-agent-skills (+2,286 ★)
plug-and-play agent skills for finance, research, science, engineering, and writing. integrates with multiple agent frameworks and supports web research, bioinformatics, cheminformatics, and analysis pipelines.
4. daily_stock_analysis (+1,272 ★)
LLM-powered stock analysis platform covering US, Hong Kong, and Chinese equities. combines market data, real-time news, AI dashboards, automated reporting, and multi-channel notifications with near-zero operating cost.
5. QuantDinger (+1,242 ★)
AI quantitative trading platform for crypto, stocks, and forex. includes live trading, strategy backtesting, market analytics, and broker integrations. built for traders experimenting with AI-assisted quant workflows.
6. Vibe-Trading (+1,148 ★)
personal AI trading agent focused on algorithmic trading and backtesting. combines lightweight automation with agent-style portfolio management and strategy experimentation.
7. FinceptTerminal (+878 ★)
modern open-source finance terminal inspired by Bloomberg-style workflows. provides market analytics, investment research, trading tools, and AI-powered financial infrastructure in one interface.
8. TradingAgents-CN (+739 ★)
Chinese-enhanced version of TradingAgents. adapts the multi-agent LLM trading framework for Chinese financial markets, datasets, and workflows. rapidly growing among Chinese quant and AI communities.
9. last30days-skill (+694 ★)
AI agent skill for researching trends across Reddit, X, YouTube, Hacker News, Polymarket, and the broader web. designed for signal discovery, narrative tracking, and internet-wide monitoring.
10. qlib (+680 ★)
Microsoft’s AI-oriented quant investment platform. covers the entire quant pipeline from data collection to alpha generation, portfolio construction, and execution. still one of the strongest open-source quant ecosystems available.
bookmark this and start today.
Anthropic tried to kill 8,100 GitHub repos. Then this happened
> They filed a DMCA. GitHub nuked the entire network within hours. Developers got notices for forks of Anthropic's OWN public repo - one guy's fork had zero leaked code.
> Boris Cherny, head of Claude Code, had to go on X personally: "This was not intentional. Should be better now."
> Meanwhile Sigrid Jin - who used 25 billion Claude Code tokens last year - woke up at 4AM and rewrote the entire thing in Python before sunrise. DMCA can't touch a clean-room rewrite.
> It hit 50K stars in 2 hours. Fastest repo in GitHub history.
> Today claw-code officially launched as an independent project with a formal press release. And the Rust port merged today - what started as a panic rewrite now ships release 0.1.0.
> 140K stars. 102K forks. More than Anthropic's own repo.
> 512,000 lines are in the wild forever. What started as Anthropic's biggest embarrassment just became their most dangerous competitor.
You cannot make this up.
Claude Code + Polymarket + Obsidian - $10K/daily trading setup ( full guide )
build your own Polymarket brain powered by Claude for daily trading
most Claude articles across X are just slop - this one will give you a real edge
result claude will analyse this in 1 promt:
• { daily } - your daily trading journal
• { markets } - individual market analysis notes
• { theses } - your big-picture trading ideas
• { research } - news, OSINT, source material
• { trades } - trade log with entry/exit
• { whales } - smart money tracking notes
• { templates } - templates for all of the above
• { commands } - your custom AI commands
when this data is collected, you’ll be able to run your custom trading commands based on it.
read the full article & set up Polymarket Claude Intel ↓
don't ask claude to predict the outcome, that's a coin flip
ask it to analyze the behavioral fingerprints of the 0.51% traders who actually make money
i just replaced a 4,000-line decision engine with 20 lines of claude API logic
the stack:
86M historical trades indexed into polars - pandas dies at this scale, polars sweeps it in seconds
filter for apex wallets: sharpe > 2.0, win rate > 65%, 1,000+ trades minimum
0.51% survive
polyterm monitors their L2 activity in real time
when apex wallets build consensus in a niche market, anomaly gets flagged instantly
then claude gets the delta:
wallet consensus + price lag + order book depth
output: {action, confidence, size_pct}
structured JSON based on contextual intent, not price points
engineering in 2026 isn't about coding rules
it's about curating the right signal for the model
if you feed claude the raw behavior of apex predators, alpha becomes a calculation
not a guess
This is how you run a zero human AI Agent Company in 2026.
OpenClaw, Cursor, and Codex agents organized under one org structure, pointed at one goal.
Get started in just one command.
100% Opensource.
Yesterday, Sunday, Benjamin Netanyahu was seen eating a choripán at a street stall near the Unión de Santa Fe stadium in Argentina. In the photo, he is seen sharing the famous "Chori Argento" with Ayatollah Jr. and an unidentified man. The photo was taken before the match between Unión de Santa Fe and Boca Juniors.
This is wild.
Someone just open-sourced a 1-person Wall Street AI agent that comes with:
- Research Desk
- Quant team
- Trading floor
- Risk Management
100% open source: