Canada 🍁 Healthcare Software Engineer | Building practical AI agents with Hermes/OpenClaw 🦞 Codex workflows, homelab notes, repo finds, and useful dev tools.github.com/benthompsonbtc Hamilton, OntarioJoined April 2026
Codex Sites looks pretty cool.
Turning a plan or dashboard into something shareable right from Codex seems useful.
Need to try it, but cool release.
openai.com/index/codex-fo…
Hermes Desktop is in public preview now. Pretty cool update from Nous.
If you like IT tools, automation, homelab setups, or AI agents, this seems like a fun one to try. Less setup friction, more room to just open it up and explore.
hermes-agent.nousresearch.com/desktop
MiniMax M3 is out. If you’re using MiniMax in Hermes Agent or OpenClaw, upgrade from MiniMax-M2.7 to MiniMax-M3 and try it out.
Looks like a big update for coding agents, long context, and tool-heavy workflows.
minimax.io/models/text/m3
Found this self-hosted agent memory/MCP project.
The part I’d test: can it remember project rules across Codex/Claude sessions without stuffing the prompt each time?
If it works, that’s a real quality-of-life upgrade for coding agents.
github.com/vectorize-io/h…
@maheshnani122 Better usage limits, better output, and better integration with opensource tools like Hermes and Openclaw 🦞 Much more consumer friendly currently than Claude is.
This is the Codex update I was waiting for.
Not because it looks flashy, but because Windows is where a lot of real setups live.
My first test: can it do boring tasks without making a mess while I steer from mobile?
A win for us non-Mac plebs.
x.com/openai/status/…
Windows users, this one’s for you.
Computer use now works on Windows, so Codex can take action on your Windows computer.
And with Windows support for Codex in the ChatGPT mobile app, you can start, review, and steer tasks on the go while work continues on your Windows machine.
Copilot Memory getting a repo-level off switch and CLI controls is boring but important.
I care less about agents remembering more and more about seeing the scope and deleting stale stuff.
Memory without cleanup turns into fake certainty pretty fast.
github.blog/changelog/2026…
@cline Any fresh models for frontier agents like Google, OpenAi and Anthropic have negligable differences, don’t buy into the hype. IMO what matters is how easy those models integrate with the tools you use, the usage limits for each and how well they integrate within your workspace.
Opus 4.8 looks worth testing for one practical reason: Anthropic says it’s better at flagging uncertainty and catching flaws in its own code.
For coding agents: less “done ✅” energy, more “here’s what might be wrong.”
anthropic.com/news/claude-op…
@nytimes If only they had the compute / usage limits to live up to that valuation. Biggest issue I’ve seen so far using both Anthropic and OpenA but Anthropic much more so.
1 Followers 33 FollowingLocal-first code intel for AI agents. MCP server, symbol graph, blast-radius. 43× fewer tokens than grep on the public bench. MIT, zero-config.
2 Followers 11 FollowingWe ship the trading machine the cloud can't touch. Local AI crypto trading + on-prem AI for regulated industries. 0 bytes leaving your hardware.
315 Followers 2K FollowingIndependent free thinker and mostly nice person. Small biz owner for life running a new SaaS startup. Here to learn and have fun.
0 Followers 68 FollowingTactical AI journaling system. Tracks clarity, discipline, and daily action on a 1–1000 scale. Building tools to replace drift with grounded momentum.
69K Followers 70 FollowingAI Edge by @milesdeutscher ⚡ | Giving you the edge on AI. Breaking news • Practical guides • Smart insights • Tips & more | Your go-to hub for everything AI.
70 Followers 62 Following25 yrs shipping software. Founder @ozthra_ai. CTO @Yorosis. Claude + Codex + Cursor/Grok daily. Honest writing on what AI actually does to engineering teams.
209K Followers 3K FollowingFollow for posts about GitHub repos, DSPy, and agents
Subscribe for top posts
DM to share your AI project (Due to volume of DMs I'll prioritize subscribers)
138 Followers 833 FollowingBuilding OpenTeam and Trending8. Build your AI agent team with OpenTeam, leveraging Claude, Codex, and similar technologies in minutes.
6K Followers 565 FollowingOperational control for Kubernetes, Docker, and Podman — without the specialist overhead. For Enterprise IT and Industrial IoT
https://t.co/5VONvQS7NQ
81K Followers 1 FollowingWorkflow automation for technical teams to build AI solutions that integrate with any app or API at no-code speed and code flexibility. Open and self-hostable
8 Followers 10 FollowingMission Control dashboard for your OpenClaw agents. Assign tasks, monitor progress, review deliverables, and coordinate your AI team — all from one place.
4 Followers 9 FollowingEnterprise IT by day, homelab builder by night. Building open source tools around networking, automation and self-hosting. Creator of LanLens.
269 Followers 403 FollowingProduct Lead @ LinkedIn Trust, xMeta head of GenAI Red Team. Exploring AI, Agents, AI Societal Impact, AI Trust, Safety & Security.
14 Followers 0 FollowingBuilding software, agents, and better workflows. Writing about devtools, infra, security, and the second-order effects of big-tech shifts. No hype, just useful
2K Followers 2K FollowingOrchestrating the next era of AI agents ⚙️ | @ApacheAirflow Committer & PMC member | Sr Eng Director @ @astronomerio | Open-source advocate
29K Followers 5K FollowingCo-founded June ("self-driving oven" acquired by @webergrills) & the co that became @Lyft. Building again, more soon. OS: @slashlast30days 26k★ @ppressdev 4.1k★
158K Followers 1K FollowingBuilding AI that upgrades humans and companies.
- PAI: https://t.co/16YCTsCgOu
- Human 3.0: https://t.co/PpczU49ANh
- Surface: https://t.co/mDwZirfm0A