Mark C AI | Builder @RealMarkCai

Holy sh*t, I just turned my TV into Jarvis with Opus 5.5 + ChatGPT voice mode I reverse-engineered my Samsung TV with Opus 5.5... 1 sentence out loud = 6 Claude Code agents opening on my living room TV. here's the full guide step by step: step 1 → put your Samsung Smart TV and your laptop on the same Wi-Fi step 2 → paste the prompt below into Claude Code on Opus 5.5 with a photo of Settings → Support → About This TV step 3 → Opus maps how the TV takes orders and finds the loophole: Samsung's own remote-control API on your home network can open any web page in the TV browser step 4 → it spins up 6 Claude Code agents in a tmux grid on the laptop and streams them as a private page on your Wi-Fi only step 5 → it builds a tiny "Jarvis" connector with one single tool: open the agent wall. ChatGPT voice mode is the ears, the connector is the hands step 6 → press Allow on the TV once, the last time you touch the remote step 7 → say "Open Claude Code CLI on my TV" and watch the windows pop in one by one the result: the laptop does the work, the TV is the control room... p.s: works on most Samsung Smart TVs with Tizen. mine in the video runs Smart Hub. The remote now lives in a drawer Find full prompt below Then read the article below on how to reverse-engineer anything

Movez @0xMovez

a week ago

x.com/i/article/2109…

10 87 702 566K 2K

And.... There it is. Claude's response to the Dots / Muse / Grok Bots of the world. We knew it was coming, just when. And what it will include. I don't have access to it yet, but I'm looking forward to diving deep into once I get access.

Claude @claudeai

19 hours ago

We’re opening a research preview of Claude Duet, a new app where Claude works as a proactive agent. It comes to you with what needs your attention, completes tasks you ask of it, keeps working when you step away, and checks with you before it acts on your behalf.

2K 4K 50K 7.4M 16K

we just shipped the last @Blender AI plugin you’ll install > 500 models in one tab. image, video, 3D, audio, textures, worlds > for humans or agents > open source > live now

🚨THIS JUST KILLED MUSE, GROK BOT AND INSTINCT IN ONE GO It's called Fo MCP and it plugs straight into the AI agent you already use. It allows you to: → Make real phone calls from your terminal → Book flights, tables and appointments → Find a time and send the calendar invite → Shop a full look from one photo → Control your lamp, smart bulbs and robot vacuum No new app. No switching tabs. Your agent plans it. Fo does it. Every AI agent in the world can now act in the real world. This is how one person builds an entire business. 🔖 Save it.

Shivani Poddar @shivanipod

3 days ago

AI can think. AI can talk. AI can get things done. Today, we're giving Fo a way to feel and control the physical world. - a head shake and a reminder in a lamp to get you going - nudge your vacuum cleaner for a deep clean - make your cli call you for updates Introducing Fo MCP.

118 107 1K 1.1M 1K

@enzo66lin Fair point, a forgotten stash is just a slower way to lose work. Flagging the line and leaving the WIP alone is probably the better default. Rollback only makes sense if it tells you it happened.

@thierrybijou Agreed. When my 9 agents hit token limits mid-task, anything not written down was just gone. Do you keep one plan file per task, or one running log?

@luqman_talks I think the line is understanding, not who typed it. AI-written code you've read and tested holds up fine. Code you never looked at is the part that collapses.

@VicNexAI First the error code from the processor, then whether Mastercard is actually enabled on the account and the currency or 3DS settings. Good test of whether someone reads logs or just reprompts the AI.

@rahozel Building in public with AI is weirdly fun because the failures are content too. Sharing what broke usually gets better conversations than sharing what worked.

@imsalariasahil The "handle work traditionally done by people" part is the real lesson. Agents get valuable when they replace a clear, boring job, not when they do something flashy once.

@suraj_sharma14 Love that tiny teams are getting backed as a category now. Credits help, but for most solo builders the hard part isn't compute, it's finding the first people who'll actually pay.

@omarvvvr I wouldn't try. If it works daily, switching costs more than it saves. Better question: what's the one task where Claude Code keeps letting you down? That's where a second tool earns its spot.

@alessandrorisk @AnthropicAI @deepseek_ai @EpochAIResearch Agreed on low lock-in. When models are interchangeable, the value moves to whoever owns the workflow and the customer.

@charliejhills Neat for keeping momentum. My worry would be quality drift when a task switches models halfway through. Does it keep the same context, or does the new model start half blind?

@thisguyknowsai Context bloat is the silent killer. The worst part isn't the limit, it's the agent forgetting which files it was editing after compaction.

@RIslam775 Without memory, every session starts from zero and repeats the same mistakes. The hard part is deciding what's worth remembering versus what's noise.

@0xrux Compact page snapshots are the quietly important part. Raw page dumps eat context fast. Curious how it holds up on sites that change their layout every week.

@smallbiztrends Sharing a workflow across teams is smart. Most small businesses don't need a custom agent, they need the one good workflow someone already figured out, copied and reused.

@BIGHEM_ @termix_ai The interesting part is trust. Humans lean on reviews and reputation. What does a track record even look like for an agent hiring another agent? Probably logged outcomes, not star ratings.

@_realhardik Least privilege for agents should be the default, not a nice to have. Easy to give one full access during a demo and then forget to pull it back when it goes live.