Ateeq (@TPateeq) header
Ateeq (@TPateeq) profile picture

Founding Engineer - @graphify (YC S26) Building https://teapawt.org/t.co/OYfHCQHW84 @ycombinator reject (3x) learning as much as I can by breaking stuff

SF / HYD
Joined August 2017
Photo by Ateeq (@TPateeq)Photo by Ateeq (@TPateeq)Photo by Ateeq (@TPateeq)Photo by Ateeq (@TPateeq)Photo by Ateeq (@TPateeq)Photo by Ateeq (@TPateeq)Photo by Ateeq (@TPateeq)Photo by Ateeq (@TPateeq)Photo by Ateeq (@TPateeq)Photo by Ateeq (@TPateeq)
The highest I have seen is a singular claude code session taking 2.5 gigs of RAM just because I was pulling it for a few weeks. 10gigs is wild
Guys, plz don't let this become an IQ test.. These memory usage comes from AGENTS running inside of Ghostty or Orca. (e.g. a single Claude Code / Next.js server is a few GB already) We help you run hundreds of agents / dev servers. And these things cost memory! Not Ghostty or Orca itself.
2
4
298
Ateeq retweeted
"R/whoosh" is internet slang (from a Reddit community) for when a joke or sarcasm completely flies over someone's head. It means "you missed the joke" — the humor whooshed right past you without landing.
1
1
46
shipping every day like my life depends on it
2
1
25
1,673
Why are MCP tool schemas flat? You could show the whole tree upfront, names only, and defer the schemas until something actually gets picked. You'd get the semantic grouping for free and spend fewer tokens doing it. Why is this not more logical....
2
4
306
🥁 announcing the first ever *Ramadan Global Buildathon* Build.LaunchGood.com Dec 7-11 $$$ in prizes ps - many cool surprises in store, stay tuned!
19
20
5
116
7,518
first page I check every morning
8
377
scary stuff
1
7
330
Ateeq retweeted
today we're launching offense: the security toolbox for coding agents. offense.sh your agent already writes the code, runs the tests, and opens the PR. offense lets it do the security work in the same loop, before anything ships. what it can run: > on every PR: dependency audits, source checks, secret scanning, github actions review > on your infra: containers and IaC > on a running app: API testing and fuzzing from your openapi spec, injection checks, exposure checks > when you want depth: a pen test agent in a disposable sandbox, with your app running inside it opengrep, gitleaks, trivy, schemathesis, ZAP, nuclei and more, behind one CLI. everything comes back as evidence your agent can read, fix from, and retest. every report also says what was not tested. built on top of boat.dev hosted starts at $20/mo. pilot access is open -> offense.sh
54
39
9
347
22,279
One software that does everything, thats the goal it seems.
Starting to roll out today: Claude Whiteboards, Claude Sheets and Claude Data. Think a plan through on a board, build a tracker in a sheet, and tell a story with your data.
5
367
Introducing patch: build and play games. - Every game dev tool in one CLI and MCP - 50,000+ open-source assets - Remix any game on patch, or mod the ones on your PC - Ship it to everyone on patch, opensource :) We're opening it to everyone today: patch.gg
116
205
53
2,335
507,603
Bro is single-handedly carrying the entire GTM for Meta
pov: training everyone’s muse to dance before our launch this week
12
9
721
71,260
Bars
This is @paulg in 2007..... This would give Shakespeare a run for his money if he were alive
2
282
Can't say nothing more
One sign up on graphify.com every minute of the day, we are nearly there We are growing more than 50% week over week, doing 5M MCP calls per month (yes agents love graphify) with over 42,000 users in 9 weeks
5
355
Everyone is arguing about AI sandbox pricing, but they're not even comparing the same thing. @boatdotdev sells a whole VM at a flat hourly rate, $0.018 an hour, so you can predict the bill. @Agent37HQ sells minutes and sleeps the box down to just disk, so an agent that mostly sits idle costs less. @vercel Sandbox only bills CPU while your code is actually running, which makes sense because an agent spends most of its life waiting on an LLM response. @e2b sells the trust layer, a real microVM per session, for running code you do not trust. @daytonaio sells pause and fork, so you can freeze a running machine and clone it into twenty copies. And there are about 30 of these on one public benchmark alone, which probably makes it the most competitive space in infra right now. Feels a lot like the vector DB rush in 2023, when everyone shipped roughly the same thing and then spent a year posting benchmarks at each other 😂
24
6
10
141
20,342
Good morning
4
348
People always ask me why I studied mechanical engineering if I wanted to end up working in software. How do I explain that I was just 3 steps ahead?
started learning to weld, just in case
3
17
569
mark my words. in 5 years no one will remember the AI giants tried personal finance. it's google+ all over again. big launch. people will try it. then the labs will realize it’s not their core business, just a distraction. and they’ll kill it. @treasury_sh will be the one still standing.
Introducing Personal Finance in Claude, in beta in the US. Connect the accounts you choose, and Claude answers your money questions in everyday language, from your own balances and transactions. Access is read-only, and Claude can't move, manage, or invest your money.
1
3
1
7
631
Holy Moly Ratatouille
sorry @lovable, here are not 100 but 1500 agents playing minecraft on boat.dev all in game in <30s from a snapshot 1500 VMs, 6k vCPU, 12TB of ram, 75TB nvme snapshotted live just $0.015/s for all this so, lovable, moving to boat.dev already?
1
4
284
Just the thing I needed
Introducing Compound Security An open-source skill that helps your coding agent find vulnerabilities, patch them, and test that your app still works. Here’s how you use it: → Help: figure out where to start → Brainstorm: pick what matters → Patch: plan, fix, and verify → LFG: take it through the workflow and more :)
1
1
3
278