The R&D Guy. Used to know about JavaScript, Docker and some stuff, now AI took over.
- Tweets 1,830
- Following 549
- Followers 1,107
- Likes 2,774
Claude is exhausting to argue. It cannot fill this, it cannot fill that. It keeps arguing back.
2
96
Canceled $1200+ worth of yearly subscriptions. @claudeai and @automatio_ai gave me some good amount of credits to move forward.
8
202
DeepSeek came a long way since the beginning. The new flash model is comparable to Sonnet/Haiku in multiple benchmarks at a fraction of cost.
1
3
176
People are bad-mouthing @finnbuilt from Bangladesh, for being a Bangladeshi and creating a profile similar to @marclou, it's alright to follow his style if its American, but not if its Bangladeshi?
I have nothing against the claims, except some of the claims are also bs themselves. You can clearly see the images are picked up from google when making a false claim.
What's wrong with people? Seriously!
4
1
16
708
The amount of unfair racism you face as a Bangladeshi from random white guys on Twitter is insane. Do they think the world belongs to them? No, the world doesn't belong to a specific race, kind or people.
4
171
Thank you. Future goes to the developers with a bit more Cyber Security knowledge.
❗️ The CEO of a South African coffee chain vibe-coded an internal dashboard for all of his 132 stores, only it had one big flaw: all of the sensitive internal data was stored in the source code of the login screen.
The dashboard URL was discovered through employees posting photos of their laptop screens on Instagram.
The CEO said he had discovered Claude and was "almost certain" that software itself will soon "become free," they said.
1
2
126
Mine looks fine I guess.
how to save your x account
x now shows you the labels that limit your reaching
> first step
- go to nitter.siddz.com/i/under_the_hood
- you need account 1+ year old + 10 posts last month
- tap view report
> what the report shows
- post labels: how many posts got flagged
- account labels: days the whole account was limited
- spam high recall = that post hidden from non-followers
- account level spam high recall is worse (hits all posts)
> finding the exact post
- report only shows the count, not which post
- download the json from under the hood
- download your full archive: settings > your account > download an archive of your data > request archive
- wait for the email, download the zip, open tweets.js
- drop both in grok and ask: "rank the most likely posts that triggered spam high recall and why"
> other ways to check
- open your posts > scroll replies > tap "show probable spam"
- settings > your account > account status
- scan recent followers for bots (default avatars, spam bios)
> things you need to know
- report covers previous month only
- 1 - 2 posts flagged out of hundreds is normal
- high recall = you need to focus more on creating positive original contents
- account labels matter more than one post flag
> what not to do
- don't panic
- don't buy third party "spam fix" tools
- don't mass delete posts
- don't spam a lot & lose your account
bookmark this. check yours now
5
234
Be careful out there
Why so many paying Claude Pro & Max ($200) developers are waking up to the “Your account will remain closed” screen right now:
Anthropic banned 11.4M accounts in H1 2026 (up 7.9×), and 89.45% of appeals (356K of 398K) get permanently rejected.
The 5 automated triggers catching normal developers:
1️⃣ Third-party OAuth → Logging into OpenCode, Roo Code, Cline, or OpenClaw with a Pro/Max subscription instead of an API key
2️⃣ VPN & IP hops → Switching VPN nodes, using Hetzner/VPS datacenter IPs, or traveling abroad
3️⃣ Billing changes → Upgrading 5x → 20x Max or switching cards (GitHub Issue #5088)
4️⃣ Automated cron/bots → Calling consumer plans via scripts
5️⃣ Security prompts → Running cyber/red-team code without enterprise clearance
Export your local `.md` context regularly once this screen hits, there is no second appeal.
2
3
284
Bring your own sandbox. (BYOS) -_-
🏖️ How to run your own local sandboxes (give this to your agent):
A spare Linux box can run your coding agent's sandboxes with each one getting its own microVM and private env vars
Mine's running on a Framework, completely replaced Vercel Sandbox for my demos
- @microsandbox runs each sandbox as a VM with its own kernel instead of a container
- the VM gets a placeholder key; the proxy swaps in the real one only on calls to that host and keys expire with the sandbox
- a firewall keeps it off the host's own services, my home network and tailnet; only the public internet
- each one gets 2 CPUs and 4 GB, up to 10 run at once and stops after idle for 30 minutes
prompt in reply
one line install:
curl -fsSL install.microsandbox.dev | sh
2
153
You can use Argon from Antigravity in other harness for sure using various techniques, but you can be sure your Google account will be marked as "Are Gone!". Real AGI.
1
121
I've just fired a client and I'm not disappointed. We thought the solution was to build a prediction model from existing data. Opus, Astra all thought the same.
He gave us a project but when asked for his collaboration and help solving the core problem he is trying to solve, he insisted on a very specific magical solution that doesn't exist ethically anyhow.
We kept asking about things, like can you forward this and that mail to us? He insisted we just have to do this and that, ofc the method he is describing doesn't exist in the most legal way, or even a gray way.
He kept mentioning about some other person who were successfully able to solve it. From our research, that guy also did something similar.
After a long back and forth,I finally snapped and said I am unable to continue if he cannot collaborate. I told him to go back to whoever solved the problem.
I understand I might not be the best person to work with if you keep telling me how to do my work, and if you do not collaborate with me when I ask for something.
If someone else can already solve the issue they way you want, why even come to me?
If you come to me, you have to trust me.
5
17
707
