FormatC:

@RyanBaileyEdge @openclaw lol, i stopped updating at 4.23, never again 24 hours of debugging and trying to fix the system, everything works as i need right now. no need to update

this crashed my system.. and removed the CODEX Support please help Here's a short version for support: Bug report: Codex plugin fails to load on Windows due to path validator + username with space OpenClaw 5.18 on Windows 11. Username folder is C:\Users\Ryan Bailey\ (contains a space). After upgrading from 4.9 to 5.18, the externalized Codex plugin (@openclaw/codex) fails to load with: runtime extension entry failed plugin directory boundary checks: ./dist/index.js The plugin files are all present and valid at C:\Users\Ryan Bailey\.openclaw\npm\node_modules\@openclaw\codex\, including dist/index.js. The manifest at openclaw.plugin.json is well-formed. Discord and Brave plugins (also externalized in 5.18) load fine — only the Codex harness fails. My read is that the harness registration path runs stricter directory boundary validation than the channel/tool plugin paths, and somewhere in that validation the space in the path causes a mismatch between expected and resolved paths (URL encoding, normalization, or path comparison logic). Workarounds attempted, all failed: openclaw plugins install @openclaw/codex --force Adding manifestPath to the plugin config entry (schema rejects unknown property) Using the Windows 8.3 short path (C:\Users\RYANBA~1\...) in config Impact: Without Codex registered, openai/gpt-5.5 can't route through its intended harness. Falls back to Anthropic CLI or xAI API, which significantly raises cost for high-volume workloads (mine is ~7M tokens/day). Suggested fix: Path normalization in the harness boundary validator should handle spaces in Windows paths consistently. Either decode URL-encoded paths before comparison, or compare resolved absolute paths via path.resolve() on both sides. Happy to provide full gateway logs and config if useful.

OpenClaw 2026.5.18 is live 🤖 xAI/Grok OAuth + sidecar auth fixes 🎙️ Realtime Android Talk Mode 💬 Telegram media + forum-topic delivery fixes 🪟 Browser dialogs visible + answerable A week of polish, plumbing, and fewer papercuts. github.com/openclaw/openc…

@jswihart @zodl_app Hardware wallet signing as well?

We’re adding Zcash coinholder polling in @zodl_app with the first poll targeted for June. Soon, anyone using Zodl will be able to share their opinion on the direction of the Zcash protocol by voting their coins. The voting protocol will also be available for integration with other shielded ZEC wallets. The draft questions and timetable for the Zcash Network Upgrade 7 (NU7) poll are available here: forum.zcashcommunity.com/t/nu7-sentimen… Thanks to @zkDragon and his team for all their work building and helping us integrate the new voting protocol.

@MinaProtocol Who? Invest in some publicity

Mesa Trail is a GO 🎉 ✅Code freeze complete ✅ Release candidate 4.0.0-rc1 tagged ✅ Preflight Mesa Testnet upgraded ✅ 12 dry runs completed ✅ Load & cluster testing succesful After months of rigorous testing, we're ready to kick off the next phase in the Road to Mesa: MESA TRAIL! Mesa Trail is the final testnet for the Mesa upgrade, designed to simulate the exact hard fork process —running from a Berkeley state to Mesa.

@YourFriendAndy @umbrel @Public_Pool_BTC my bitaxe worked great for 3 weeks. then for no reason it stopped working forever. Support says it is what it is and if i want i can buy a new one with a small discount. ridiculous.

Simple Bitcoin hardware stack: -PC running @umbrel OS -Bitcoin node on Umbrel -@Public_Pool_BTC on Umbrel -Small solo miner like a Bitaxe With this, you run the entire Bitcoin blockchain (node), your own mining pool (Public Pool), and miner (Bitaxe)... All from home. Locally.

@IBI_Investment ואו מתאים בול לקניות חצי שבועיות בסופר!

הצטרפו למסחר בבורסה עם IBI SMART - קבלו 300 ₪ מתנה לחשבון.

@Emarged Aliexpress version within 3...2...1..

This is exactly how a multi-billion-dollar company is created. Fortune always favours those who solve day-to-day problems.

@PaulRevere3MTA3 @gainzy222 sounds like a good idea, thanks Paul

@gainzy222 How’s Zcash going bro? Can I still put it on a ledger and pay for my kids college in a few years? 🤡

This period is as close to pre-2017 as I can think of There’s no “consensus” play that won’t run because everyone is too poor and under leveraged, risk appetite lost The new culture of “hodl” is ingrained because all we can do is hold & dca, we have no money to speculate/gamble

@venturetwins If only any of the main characters stays and looked the same.

He’s a true creative talent - both ideas and execution. Another one 👇 instagram.com/reel/DXOwuZ9k7…

I am obsessed with this Japanese man using AI video to put himself into movies (he's on IG at @ai_am_furufuru)

@zedbtc21 @openclaw lol same, never updating again

@openclaw After days of trying to fix, just to have to revert back to 4.23 I'm surely not going to update until there is agreement this version actually works.

OpenClaw 2026.5.2 🦞 🧠 xAI Grok 4.3 🔌 Plugin installs/updates are sturdier ⚡ Gateway + agent hot paths are leaner 💬 Discord, Slack, Telegram, WhatsApp fixes 🎙️ TTS, Realtime, web search, voice-call polish Less drama. More uptime. github.com/openclaw/openc…

@openclaw As per usual, this broke everything. Moving to hermes

OpenClaw 2026.4.29 🦞 💬 Group chats feel much better now 📌 Follow-up commitments from context 🔐 Safer exec, pairing, and owner controls 🟩 NVIDIA provider + model catalogs ⚡ Faster startup + plugin/channel fixes Group chat finally feels agent-native. github.com/openclaw/openc…

@bottomrightside @openclaw Same here, so its broken. cant get gateway up.

@openclaw Its wild that I can send a non updated agent to fix the updated agent that broke from the update. bonjour/mDNS crash loop

OpenClaw 2026.4.24 🦞 ☎️ Voice calls can now reach the full agent 🧠 DeepSeek V4 Flash + Pro join the team 🖱️ Browser automation got coordinate clicks + better recovery 🔧 Telegram, Slack, MCP, sessions, and TTS fixes More reach. Less duct tape. github.com/openclaw/openc…

@AvivitGutman @Cursedfiona לא שמעת על ״פרידמן״ שהיה מילת גנאי לאשכנזים בתקופת בובליל והאח הגדול...?

@Cursedfiona חייבת למחות נגד השימוש בשם "נהוראים". עוול עצום לאלפי אנשים, נערים וילדים על לא עוול בכפם. גועל נפש.

הרצח בפתח תקווה קרה ברחוב שלי. זה לא מה שמזעזע אותי, כי אני יודעת איזה נהוראים יש כאן. מפחדת להעיר להם ואומרת לבעלי גם לא להתעסק איתם.

@barzik @turgeman_reut Hi Ran what bash do

@turgeman_reut מי שיצא עלייך בתגובות מעולם לא היה בפוזיציה שהוא מקבל הרבה בקשות/דרישות/פניות. נכון, יש את העולם האידיאלי בו יש לכל אחד קשב בלתי מוגבל אבל המציאות היא לא שיעור חברה של כיתה ו׳. בפועל יש לאנשים עם המון חיים 24 שעות ומעט מאד זמן. את עוד היית נדיבה. במקרים כאלו אני אפילו לא עונה.

הרבה הקציפו בתגובות, אני מעבירה רגע את הפוקוס מהתגובה שלי, כי זו נקודה חשובה- כשאתם פונים לאנשים מיוזמתכם עדיף לתת ערך/קונטקסט על הone. לא משנה באיזה פוזיציה אתם, יזמים, סטודנטים או מנהלי קהילה- אל תצפו שהצד השני ״יחלוב״ אתכם עם שאלות מגששות. למדתי את זה בדרך הקשה(עד מאוד)

@robustus @konstructivizm Youre helping him grow, i wouldnt see it if u did t comment

@konstructivizm Another fake shot. You just trolling, or are fake visuals your whole engagement strategy?

Artemis II Orion Capsule Moments Away from Ocean Splashdown

@o1_labs @MinaProtocol Whats mina

We've just released a new open-source Mina explorer with complete block and transaction history across all networks! 🔍 Built for our Mesa development work, and now available as another option for the community to explore network activity 🙌 👉 o1-labs.github.io/mina-explorer

@gainzy222 Move to an apartment with mamad and sit there on your computer all night uninterrupted

well played, right when i was falling back asleep

daily past midnight rocket volley to wake u 👍

@samuraipreneur Context window? Any limits? Has to be 4k-8k?

okey, this is crazy nuts. you can now run super smart intelligence on literally ANY device. - on your phone - on your laptop - even on a raspberry pi fully open source. completely free. the 9B model performs on par with GPT-OSS 120B while being 13x smaller. it beats models 4x its size. and delivers performance that’s competitive with sonnet 4.5 and gemini 3 flash intelligence just went local, smaller, faster and smarter. open source always wins.

@Axel_bitblaze69 What about the large context window needed by openclaw and all its files? (128k+)

ok sharing how to actually use qwen 3.5 models, first what you need is, > decent laptop/desktop (M1/M2/M3 Mac or windows/linux with 8GB+ RAM) > python installed > that's it step 1: install ollama (easiest way to run local models) mac: curl -fsSL https://ollama. com/install.sh | sh windows: download from ollama. com/download this is the simplest way to run models locally. no complicated setup. step 2: download qwen models open terminal and run: ollama pull qwen2.5:0.5b ollama pull qwen2.5:3b ollama pull qwen2.5:7b note: they're using qwen2.5 naming in ollama. same models, different numbering. each download takes 2-10 minutes depending on size and internet speed. step 3: test if it works ollama run qwen2.5:3b you'll get a chat interface. ask it something simple: "what is bitcoin?" if it responds, you're good to go. ctrl+d to exit. which model for what: qwen2.5:0.5b - super fast, basic checks • "is btc above $60k?" • "did this transaction happen?" • use for: alerts, simple yes/no questions qwen2.5:3b - balanced, most versatile • wallet monitoring • price analysis • market summaries • use for: 80% of your monitoring tasks qwen2.5:7b - deeper analysis • token comparisons • trend analysis • research summaries • use for: when you need actual thinking alternative: using with clawdbot if you're using clawdbot/openclaw framework: clawdbot config --model ollama/qwen2.5:3b then your agent uses local qwen instead of cloud APIs.

@TukiFromKL Contex window can do 128-256k on the same computer?

Do you understand what just happened? Qwen dropped small models you can run on a $600 Mac Mini. Locally. No internet. No subscription. No company controls your access. Go do this right now: Download LM Studio Search Qwen 3.5 Grab the MLX versions Load them You now have unlimited AI on your own machine. Nobody can take it away from you. Not a company. Not a government. Not a terms of service update. Everyone's fighting over who controls AI. The answer just became: you do.

@JtBlkbox @ChrisDunnTV no need. im running one perfectly on RasPi4

@ChrisDunnTV Nice, I’m about to buy a Mac mini myself to get some shit setup 🤙

I genuinely don’t get how people are playing video games now... because these ai agents just turned real life into the ultimate sandbox. You can build insane stuff, for almost nothing, in no time. Why grind fake coins when you can stack real ones faster & easier than ever?