Tim Boone

@mitchellh I feel attacked

Nobody should be using up arrows to get previous commands in a terminal or shell. You have to move your hand and its linear complexity in keystrokes. Use ctrl+p (for low n) or ctrl+r. Use a real shell or history manager (fish, fzf, atuin) for ctrl+r.

@GodsBurnt @elonmusk @nikitabier it’s posts like this that make this app incredibly difficult to trust Natl Inquirer at the grocery store vibes Muting does nothing, they return like zombies

Just got off a 3 hour call with my insider from Washington DC I can't say much right now but January 29th will start the largest transfer of wealth in human history Most people have no idea whats about to happen...

X algo has taken echo chambering to an entirely new level

@ILoveBlueBell Was I the only one who thought this was a monthly ice cream subscription offering?

𝘾𝙤𝙢𝙞𝙣𝙜 𝙎𝙤𝙤𝙣! It’s more than just protein — it’s Blue Bell PRO. Introducing NEW high-protein pints in five delicious flavors: Chocolate, Coffee, Salted Caramel, Strawberry, and Vanilla. ✔️33-35g of protein per pint ✔️0g added sugars ✔️Low fat ✔️Gluten-free

Catching up on the Greenland story Can someone give me the best TLDR?

And just like that, all the Romo haters were silenced

@rauchg @andrewqu Nice work. Was just discussing this last week x.com/vtimmyb/status…

We're introducing 𝚜𝚔𝚒𝚕𝚕𝚜 – the "npm" of AI skills. Excited to see an open, agent-agnostic ecosystem of skills flourish. To get started, try: ▲ ~/ npx skills i vercel-labs/agent-skills

Vibe app support centers will be all the rage

@davidhoang @claudeai @obsdmd @raycast Give us a try before you go through all that effort. @TryCadence

The holy trinity for the personal OS/Chief of Staff I'm building for myself: @claudeai, @obsdmd, and @raycast.

Cadence is coming. Tell it your goals. It handles the rest. —> cadence.bot

Been working on this for a while. Finally ready to share.   @TryCadence - AI goal coordination for people who actually want to get things done.   Waitlist is open → cadence.bot

@mcuban Quantity over quality

Explain to me why the insurance company will pay $2500 for an MRI when there is a center down the street that will do it for $350 ?

@RaghuRaghuram 🚀

'giving people a shot is a profound accomplishment for society and humanity. It is the gift that keeps on giving.' Lets go!!

@datarade Sorry to hear that Kumar. I know you’ll keep it going strong in his honor.

Cofounder & friend in early-stage biz died ~2mo ago He died in car crash on late-night drive returning from vacation—DOI. Was waiting to catch up with him after his vacation, shouldn't have done that. I kick myself. His last msg to me: asking if we had any 2026 customers. Met large co today who wants to be 1st customer. 😢

@ryancarson 💯 If anyone needs help doing this, we specialize in it @alt_tab_tech

If your app has an agent (and if it doesn’t, you’re in trouble) it needs to have tools to do all CRUD operations the user is able to do through the UI. 100% of them. And the agent should understand what page the user is looking at and how it works.

@snyksec @vtahowe 💯

@vTimmyB @vtahowe The problem remaining with LLMs, whether SKILLS, RULES or otherwise is that they are fundamentally unreliable and non-deterministic - hence code security SAST that relies on Symbolic AI, deep security expertise and other traits is the great companion to coding agents.

everyone seems to really be feeling the AGI lately which is exciting but I have some questions > We're now opting to generate code instead of checking out OSS repos. The good part of a package supply chain vuln is you can patch it upstream and then everyone can download the fixed version. What happens when a model codes the same insecure code pattern into thousands of repos? There's no central and systemized patch process there. > Skills are super powerful and are powering a new era of Claude Code/Claude Agent SDK wrappers. While we're all worried about leaked env files let's not forget that leaking IP is a startup ender also. Skills get thrown into your context window and it's up to the agent harness to protect it accordingly. > Yes Ralph is cool, but in order to let him cook he needs to be able to run commands freely and download whatever packages he needs. I'm adding @SocketSecurity CLI to my agent harness to make the agent check package vuln scores before downloading. I could use an MCP server for this instead but I'm hesitant due to point 2 about protecting what's in my skill files. Agent sandbox is likely also a good idea for Ralph.

I’m a moderate. Probably lean more right than left. That said, she clearly wheel turned to avoid hitting him. Just shoot the tires, not the human. Wtf is wrong with people

Waiting for the day my gf asks who this Maria Chen person is

@mattshumer_ Not to sound too cliche here, but what problem are you looking to solve? Can you give an example?

Specifically, I’m looking for a messaging system that allows multiple agents to communicate while working on tasks together. Not a human-facing Slack-type app with agents integrated… more of an agent-first devtool that allows agents to coordinate on difficult tasks.

Is anyone building a “Slack for agents”? If so, I’d love to fund you + blow you up with my platform. I’ve kickstarted growth for companies like @GroqInc, @rork and more. Would love to do this again for the right company. If this is you, reach out.

@vtahowe Skills as intellectual property is a fascinating topic. Short term will see dozens of zero to one’s. Long term my guess is it becomes commodity engineering

@vTimmyB @snyksec I think the skills that are not SME and can be distributed to others will do so through Claude plugins