Joshua Tubbs

@tayarndt @OpenAI @embirico @sama I'm in so much agreement with this. Please make it happen.

I love the work @OpenAI you are doing with codex. Wish having plans between 20 and 200 dollars. Why not a 100 dollar plan? @embirico and @sama can we make this happen?

1+ Hermes Agent.

@AmpereSh So no more pricing per 1m tokens anymore? That's what the pricing page says. Also, I have a Claw already set up. Is there a way to migrate it over to Ampere?

Still self-hosting? Move to Ampere. → Unlimited Claude — no API bills → Browser automation with real IPs → Super memory → Auto updates, zero maintenance Stop worrying about tokens. Start shipping. ampere.sh

Introducing Unlimited AI Usage with OpenClaw. That's Ampere — managed OpenClaw hosting built for people who don't want to think about API bills. Here's what you get 🧵

@ekello Link on GitHub?

Introducing the Auto Book Builder: a game-changer in writing. Powered by 'Codex,' it ensures seamless continuity, enforces consistency, and integrates custom skills to revolutionize your writing process. #WritingTools #AuthorLife

Wow. I had to stop using Codebuff when Anthropic went ban crazy. Now i can get the power of Codebuff for free. No dealing with api keys, and no setting up subagents either. Though, I wish we could add custom agents like such for the accessibility agents repo.

@jahooma Oh dang. The same multi-agent workflow as Codebuff but free? Gonna try this.

Introducing Freebuff: the free coding agent 100% free, up to 10x as fast as Claude Code npm install -g freebuff

@steipete What about native Codex support? ACP's alright but sometimes it has trouble managing processes. GSD framework kinda works but hangs up on interactive prompts.

I have that uncommitted in a branch! needs more testing. I even added mcp support 💀

openclaw 3.13 added live chrome attach. before this, my agents couldn't see what i see. dashboards behind login walls, authenticated sessions, real browser state. i had to screenshot and paste. now: # openclaw.json { "browser": { "defaultProfile": "user" } } # chrome://inspect/#remote-debugging → enable # or launch with: chrome --remote-debugging-port=9222 enable remote debugging in chrome, and the agent connects to your actual browser. logged in. real cookies. no extension needed. tested locally. told scout to check my vercel dashboard for traffic data. it opened the tab i already had logged in, read the numbers, and wrote them into the weekly report. no headless browser. no separate login. no screenshots. it just sees what i see. the gap between "my agent can browse the web" and "my agent can browse MY web" was always authentication. this closes it. ps: requires chrome running on your local machine. vps users need to connect your machine as a node.

the fastest growing GitHub projects this month: 1. openclaw/openclaw (122K stars) your own personal AI assistant, runs 24/7 on any OS (what I use to run all my agents) 2. obra/superpowers (30.7K stars) agentic skills framework. plug-and-play tools for AI agents 3. ruvnet/RuView (30.4K stars) turns regular WiFi signals into real-time human pose detection no cameras. no sensors. just WiFi. 4. 666ghj/MiroFish (17K stars) swarm intelligence engine that predicts anything 5. moeru-ai/airi (16K stars) self-hosted AI companion with real-time voice chat runs on your own machine. you own it. 6. shanraisshan/claude-code-best-practice (11.8K stars) the best practices repo for building with Claude Code 7. badlogic/pi-mono (11.8K stars) full AI agent toolkit: CLI, unified LLM API, web UI, Slack bot 8. bytedance/deer-flow (10.4K stars) ByteDance's open-source SuperAgent. researches, codes, creates on its own 9. shareAI-lab/learn-claude-code (9K stars) build a Claude Code clone from scratch. bash is all you need. 10. p-e-w/heretic (7.6K stars) removes guardrails from any language model automatically the pace of AI right now is insane. bookmark this. next month's list will look completely different.

There's a lot of cool stuff being built around openclaw. If the stock memory feature isn't great for you, check out the qmd memory plugin! If you are annoyed that your crustacean is forgetful after compaction, give github.com/martian-engine… a try!

Yes, every OC user should do this if you use it for deve owrk.

Why does the OpenClaw update command give a damn about how clean the commit is? Update Result: ERROR Root: /home/openclaw/openclaw Reason: preflight-no-good-commit Before: 2026.3.11 T I just wannatupdate.

OpenClaw 2026.3.12 🦞 🎛️ dashboard v2 — slick new control UI ⚡ /fast mode for models 🔌 ollama/sglang/vllm → plugins (core goes on a diet) 🛡️ device tokens now ephemeral ⏰ cron + windows reliability fixes touch grass? nah, touch main 🌿github.com/openclaw/openc…

@zocomputer How does handling restarts work? The traditional openclaw gateway restart commands didn't work due to issues with systemd, so I had to make my own script that would kill the process and start it manually in the foreground. Ran out of Zo credits having the aI do that.

Holy shit... Microsoft open sourced an inference framework that runs a 100B parameter LLM on a single CPU. It's called BitNet. And it does what was supposed to be impossible. No GPU. No cloud. No $10K hardware setup. Just your laptop running a 100-billion parameter model at human reading speed. Here's how it works: Every other LLM stores weights in 32-bit or 16-bit floats. BitNet uses 1.58 bits. Weights are ternary just -1, 0, or +1. That's it. No floats. No expensive matrix math. Pure integer operations your CPU was already built for. The result: - 100B model runs on a single CPU at 5-7 tokens/second - 2.37x to 6.17x faster than llama.cpp on x86 - 82% lower energy consumption on x86 CPUs - 1.37x to 5.07x speedup on ARM (your MacBook) - Memory drops by 16-32x vs full-precision models The wildest part: Accuracy barely moves. BitNet b1.58 2B4T their flagship model was trained on 4 trillion tokens and benchmarks competitively against full-precision models of the same size. The quantization isn't destroying quality. It's just removing the bloat. What this actually means: - Run AI completely offline. Your data never leaves your machine - Deploy LLMs on phones, IoT devices, edge hardware - No more cloud API bills for inference - AI in regions with no reliable internet The model supports ARM and x86. Works on your MacBook, your Linux box, your Windows machine. 27.4K GitHub stars. 2.2K forks. Built by Microsoft Research. 100% Open Source. MIT License.

@tayarndt Hmm. I know quite a lot about Git, working on something new for accessibility agents, but I like how this course is set up. Podcast links are MIA right now though.

@Orinks33 Not sure when the recordings will be available but the documentation has everything you need to get started right now. Jump in and let us know if you have questions. community-access.org/git-going-with…

This weekend was something special. We ran a 16 hour workshop teaching people how to use Git and GitHub accessibly. From the very beginning. What a repo is. How to clone, commit, push, and collaborate. All of it done with a screen reader. Then we went further. Students learned to build their own custom agents with VS Code and GitHub Copilot. @JeffBishop built image agents that are already in the workshop repo and ready to use. People came in not knowing Git. They left feeling confident enough to contribute to open source. Thank you to @JeffBishop, @mikedoise, and every single person who made this happen. This is what accessibility education looks like when the community shows up. #a11y #GitHub

Be careful about listening to anyone who uses the term "vibe coding" when referring to any code written by agents. It means they don't actually use agents to build real things. There is a world of difference between vibe coding and agentic engineering. Pretty soon, hopefully we won't need to say "agentic ..." as it'll be assumed any professional engineer is using agents to code.

🚨 Someone just open sourced a fully autonomous AI hacker and it's terrifying. It's called Shannon. Point it at your web app, and it doesn't just scan for vulnerabilities. It actually exploits them. Real injections. Real auth bypasses. Real database exfiltrations. Not alerts. Not warnings. Actual working exploits with copy-paste proof-of-concepts. Here's what this thing does autonomously: → Reads your entire source code to plan its attack → Maps every endpoint, API route, and auth mechanism → Runs Nmap, Subfinder, and WhatWeb for deep recon → Hunts for Injection, XSS, SSRF, and broken auth in parallel → Launches real browser-based exploits to prove each vulnerability → Generates a pentester-grade report with reproducible PoCs Here's the wildest part: It follows a strict "No Exploit, No Report" policy. If it can't actually break it, it doesn't report it. Zero false positives. It pointed at OWASP Juice Shop and found 20+ critical vulnerabilities in a single run including complete auth bypass and full database exfiltration. On the XBOW Benchmark (hint-free, source-aware), it scored 96.15%. Your team ships code daily with Claude Code and Cursor. Your pentest happens once a year. That's 364 days of shipping blind. Shannon closes that gap. One command. Fully autonomous. The Red Team to your vibe-coding Blue team. Every Claude coder deserves their Shannon. 10.6K GitHub stars. 1.3K forks. Already trending. 100% Open Source. AGPL-3.0 License.

@clawdtalk What's going on? I call my bot and it's saying this is Scott and it's a test call. My bot is connected.