netrunner

Super excited to announce that after 1.5 months of working in silence with my team, we've launched Luminex.io 🟨! We started as ordinals.cheap, and now we're stepping up our game - from "ordinal" to out-of-this-world! 🌎🟧 Try our top-notch minting hub rn!

@CopilotKit rendering inline is fine but the jsonl progressive stream is what matters, you see the ui build as it generates not after

Your agent describes the UI. CopilotKit renders it inline. Ask for the weather, a live weather card shows up in chat. That's A2UI. Open-JSON-UI, MCP Apps and custom specs render the same way. Learn more about it 👇 docs.copilotkit.ai/built-in-agent…

@glider__ non custodial or is the bridge holding the keys somewhere

You can now access Solana from any EVM chain on Glider. Everything you want to own from Base, Arbitrum, Ethereum, and Solana, in one account and one place. More to choose from, nothing extra to set up. 🔗

@ai_for_success no invite needed btw, just join the beta program and you got a shot. capacity limited so its a coin flip

⚡ Kimi just launched Kimi K2.7 Code HighSpeed, a faster version of their open source coding model with speeds up to 180 tok/s and 260 tok/s on shorter contexts. Rolling out first to Beta / API & Business users due to capacity limits.

@IT_Tech_PL @kiyotaka_ai negative coinbase premium on a us deal? funny

$BTC printed 63K -> 66K on news of a US-Iran peace deal. Perp CVD is 33x spot. Leverage led this move. - Spot CVD: +61.43M - buying present but lightweight - Perp CVD: +2.08B - derivatives drove the bulk, 33x spot volume - Coinbase Premium Gap: -0.07% - US buyers not leading - Funding Rate: +0.0006% - near flat, no overheating - Order Book Depth Delta: -384.60 spot / -4.5K perp - ask pressure building Both CVDs green on the news. But perp is doing 33x the work - this is leverage-driven, not organic spot demand. Coinbase Premium went negative during the spike. US buyers didn't chase. ⚠️

@CoinAnk ict structure and smc in a sidebar, retail about to discover liquidity grabs the hard way

CoinAnk Pro Chart just got a new sidebar: Tutorial. Every pattern. Every indicator. Every structure — explained in context, right next to your chart. Candlesticks, Technical Indicators, ICT Structure, Order Flow, Volume. Browse the category, read the definition, see the chart breakdown. Then hit "Identify on chart" — and watch it find every matching pattern on your current chart automatically. 👉：coinank.com/proChart?excha…

@superai_conf mistral as platinum is a flex 👀

Every name on this screen believed in what we were building before the doors opened. Thank you to the SuperAI 2026 sponsors for making this possible.

@Peter_rock07 @variational_io @nadoHQ @bulktrade @tradexyz @risextrade @tread_fi @slx_fi @get_truenorth @ProprXYZ @novadotmarkets @pacifica_fi @CarbonTerminal @01Exchange @tradehotstuff @hellotradeapp @papertrade_xyz @extendedapp this aint all of them tho, theres no exhaustive list, just whoevers timeline you scrolled

List of all Perps DEX 2026 👀 - @variational_io - @nadoHQ - @bulktrade - @tradexyz - @risextrade - @tread_fi - @slx_fi - @get_truenorth - @ProprXYZ - @novadotmarkets - @pacifica_fi - @CarbonTerminal - @01Exchange - @tradehotstuff - @hellotradeapp - @papertrade_xyz - @extendedapp - @grvt_io - @StandX_Official - @gmtrade_xyz - @EVEDEX - @SatoriFinance - @Antarctic_Ex - @reya_xyz - @DecibelTrade - @VestExchange - @hibachi_xyz - @dango - @PhoenixTrade - @etherealdex - @BullbitDEXHQ - @liquidtrading - @miracletrade - @BullpenFi - @Dexari - @perpltrade - @popdex_ - @astros_ag - @ventuals - @Rho_Labs - @Dreamcash - @Ostium - @Bulletxyz - @GTE_XYZ Did I miss something?

@nahid_pro09 free is the weights download, the api still charges you and a local run wants like 340gb ram lol. not exactly free if you actually want to use it

you can access moonshot's new kimi k2.7 code right now, and it’s officially open-source and FREE😳 most people thought gpt-5.5-level performance was years away, but this model is already here and it's rewriting the rules of what a free assistant can build takes 1 minute to test what you get with kimi k2.7: →benchmark performance nearing gpt-5.5 levels across coding tasks →30% fewer reasoning tokens consumed compared to k2.6 →full open-source access through the kimi api and kimi code environments →high-speed 6x mode currently in development the performance breakdown (vs opus 4.8 & gpt-5.5): >mcp atlas: kimi 79.4 (beats opus 76.4) >mcp mark verified: kimi 81.1 (beats opus 76.4) >program bench: kimi 53.6 full setup guide (kimi k2.7 integration): step 1: access the platform →head over to the kimi api or kimi code dashboard to switch to the k2.7 environment step 2: optimize your tokens →leverage the 30% reduction in reasoning tokens to run more complex agentic loops for less cost step 3: integrate into your workflow >plug the new model into your current ide or client to start testing against your existing benchmarks step 4: prepare for the 6x speedup >keep your eyes on the update logs for the release of the high-speed mode to crush your latency issues already using opus 4.8? skip straight to step 3 and compare the output quality on your toughest codebase migrations this is the single biggest sign that frontier intelligence is becoming a commodity and the days of charging premium access for these models are numbered bookmark this before the free access tiers change or hit capacity

@camden_hulse @treenalabs @theresidency ive tried coding on the beach in asia, the glare kills it and you cant see a thing on screen. office aircon wins every time tbh

coding in the ocean!?

@ns123abc satya doesnt care who wins he sells the compute either way

BRO LITERALLY PREDICTED THIS

@vikas_ai_ saved it, wont watch it. classic

FULL COURSE 1 HOUR (Build & Automate Anything)

@TTrimoreau idk the 99% number keeps getting thrown around but nobody can source it lol

At this point if AI writes 99% of code, who even survives in tech??

@PrajwalTomar_ context compression is cool but i wanna see the numbers on my own traffic before bookmarking anything. depends way too much on how redundant your data is

If you're serious about open source AI, bookmark this entire list. Headroom alone will save you 90% on API costs without touching your code. Most people have no idea these tools exist.

@dominik_btc I’m using a raspberry pi 16gb ram for managing multi projects via herdr (for serious stuff) and hermes for more casual stuff. While on my computers I use Claude and Codex apps for MCP connections and stuff where I need to work on UI/UX. Waiting for the Mac Studio

@netrunner_btc What’s your set up right now bro

holy, used fable5 for 24h on medium mode doing design, planning and review directing custom gpt 5.5 low effort agents on implementation, and it seems to have solved what i tried to build the last month with opus 4.6/7/8 (ultracode) and gpt 5.5 on xhigh cant wait to see gpt 5.6 bc i wont be paying for APIs regardless how good this is. love this IPO competition

@hackinarticles claude flagging this as a TOS nightmare in 3 2 1

Automated Penetration Testing with Claude AI 🔥 Telegram: t.me/hackinarticles ✴ Twitter: x.com/hackinarticles AI is rapidly changing the cybersecurity landscape, helping penetration testers automate reconnaissance, analysis, and exploitation workflows more efficiently than ever ⚠️ 📚 What You'll Learn in This Guide 🤖 Introduction to Claude AI for Pentesting ⚙️ Setting Up AI-Assisted Security Workflows 🔍 Automated Reconnaissance & Enumeration 📋 Analyzing Scan Results with AI 🎯 Vulnerability Identification & Prioritization 🚀 Automating Penetration Testing Tasks 🛠️ Integrating Claude with Security Tools 📂 Streamlining Report Generation 🧠 AI-Powered Decision Making for Red Teams 🔄 Enhancing Productivity During Assessments 🛡️ Ethical Considerations & Responsible Usage ⚠️ Limitations of AI in Offensive Security 💡 Claude AI can assist security professionals by automating repetitive tasks, interpreting security findings, generating commands, and accelerating penetration testing workflows. Human validation remains essential to ensure accuracy and responsible use. 📖 Article: hackingarticles.in/automating-pen… #ClaudeAI #ArtificialIntelligence #Pentesting #CyberSecurity #RedTeam #EthicalHacking #InfoSec #SecurityAutomation #LLM #AITools

@Taniyatweets_ depends what for tbh, qwen for local coding stuff but kimi has been surprising me lately on longer context

be honest, which is the best open source AI model? 1. Qwen 3.5 2. DeepSeek v4 3. GLM-5 4. Kimi K2.6

@cyrilXBT another one bsing for attention

Rockstar spent $1 billion and 8 years building GTA VI. Claude Fable 5 rebuilt it in 10 hours with 4 million tokens. No studio. No payroll. No budget. The people who understand what that sentence means are already building. The ones who don't are still waiting for permission to start. The wall between you and a AAA game just hit zero.

@RoundtableSpace bro i cant believe anything you post anymore, so ill take this and call bs...

CLAUDE FABLE 5 VS OPUS 4.8 VS GEMINI 3.1 PRO VS GPT 5.5.

@TedPillows whats your invalidation level on this

Peak capitulation is coming for $BTC in Q3. A small pump may happen before that, but the bottom is still not in.

@arceyul does it handle the layout or just spit a template?

Mythos Claude es una locura tutorial de 12 minutos sobre cómo crear sitios web animados y listos para vender con Claude Fable 5