abh1sek Profile Banner
Abhisek Datta Profile
Abhisek Datta

@abh1sek

Followers
3K
Following
5K
Media
216
Statuses
3K

Author of 🚀 https://t.co/Tgbp4Dx9V5 | Building @safedepio | Software Engineer | Possibly Security Researcher | Securing Open Source Software | Nerd?

India
Joined May 2008
Don't wanna be here? Send us removal request.
@safedepio
SafeDep
6 hours
The SafeDep GitHub App is officially live on the GitHub Marketplace!🎉 Install → Scan → Relax😌 Try it out here👇 https://t.co/ETDd5zfJT9
0
2
8
@abh1sek
Abhisek Datta
2 days
Books that you should read for improving your software engineering skills. Side effect is, they will prepare you for Staff+ roles - Clean Code - Clean Architecture - The Pragmatic Programmer Once you are able to use the ideas: - Designing Data Intensive Applications For
0
1
7
@abh1sek
Abhisek Datta
3 days
Looking forward to this!
@deepigoyal
Deepinder Goyal
3 days
So many people keep asking me about Continue. What is it? What are you up to? Here you go... Continue started as a research effort two years ago, with the belief that if the human body is a system, it should also have its leverage points. The simple levers that, when adjusted,
0
0
0
@abh1sek
Abhisek Datta
8 days
Looks like there is an onslaught of AI influencers on X. AI generated persona. Similar content. Anyone else seeing this?
0
0
1
@abh1sek
Abhisek Datta
9 days
🚀
@safedepio
SafeDep
10 days
When you have the SafeDep GitHub App installed, you can just sit back, relax, and focus on building. 👉 https://t.co/VUbAaMSIhI
0
0
2
@abh1sek
Abhisek Datta
9 days
This is true. I am seeing great benefit of Claude Code (and now Cursor) plan mode. In fact I think it makes sense to check in the plan based on which a certain feature waste implemented.
@mattshumer_
Matt Shumer
10 days
I’m begging you When having Codex/CC/whatever make a large change to your codebase Have a lengthy conversation with it first, clarifying everything you can Only then should you ask it to start building
0
0
1
@abh1sek
Abhisek Datta
10 days
Who is building AI fortune cookies?
0
0
0
@abh1sek
Abhisek Datta
11 days
Next is to figure out the workflow. How to use them to boost productivity, retaining control and not compensate the coding time advantage with code review slog.
0
0
0
@abh1sek
Abhisek Datta
11 days
AI coding agents starts getting meaningful when you treat them as a developer and not a replacement to your own thinking and imagination.
1
0
1
@FFmpeg
FFmpeg
11 days
Arguably the most brilliant engineer in FFmpeg left because of this. He reverse engineered dozens of codecs by hand as a volunteer. Then security "researchers" and corporate employees came along repeatedly insisted "critical" security issues were fixed immediately waving their
@FFmpeg
FFmpeg
12 days
The maintainer of libxml2 put it very well
154
709
9K
@safedepio
SafeDep
11 days
We ask @NotebookLM to give us the perfect overview—here’s the video...
0
2
5
@abh1sek
Abhisek Datta
11 days
lazygit
@Tech_girlll
Mari
12 days
Any alternative to git?
0
0
0
@abh1sek
Abhisek Datta
11 days
Anyone using AI code review tool? Which one? Pros and cons?
2
0
4
@abh1sek
Abhisek Datta
12 days
New version of vet released with multiple bug fixes and feature updates. ➡️ Policy language revamp ➡️ Multiple bug fixes ➡️ Console experience improvements Everything you need to audit, analyse and secure your open source software supply chain. https://t.co/SZavyaJRsO
Tweet card summary image
github.com
Protect against malicious open source packages 🤖. Contribute to safedep/vet development by creating an account on GitHub.
0
3
7
@abh1sek
Abhisek Datta
12 days
In another news, I have stopped using GitHub Copilot in my nvim. Its a zen experience without the zen mode. Works amazingly well when I am laying down the LLD aspects of the project. Claude Code for analysis, planning and coding few stuff as per the declared conventions.
0
0
1
@abh1sek
Abhisek Datta
12 days
Attack surface reduction ftw!
@IceSolst
solst/ICE of Astarte
13 days
Security noob mistake #1: Never say your code is “safe”
0
0
1
@abh1sek
Abhisek Datta
13 days
Fairly sure it is. Given the cost of storage and cost of engineering bandwidth for fixing security bugs.
1
0
2
@abh1sek
Abhisek Datta
13 days
In security, sometimes we overlap quantitative and qualitative solution. This is a mistake. Even with AI, it will be nearly impossible for a static code analysis tool to beat a security researcher like duke or taviso on novelty (quality). But it can beat them on quantity.
1
1
3