Senior Software Engineer.| Fitness | Living in my head all the time.
Joined May 2017
- Tweets 763
- Following 538
- Followers 85
- Likes 6,179
119 Photos and videos
Pinned Tweet
5 Apr 2023
"Hey! Check out this cool Github repo github.com/Dmoment/gpt-commi… I created to generate better commit messages using OpenAI's GPT-3. It's a game-changer for anyone who wants to streamline their Git workflow. #github #git #openai #gpt3 #commitmessages"
1
1
14
1,289
Mar 31
Spent some time setting up Claude Code for my Rails React project. Commands, rules, skills, hooks - figured out what actually works after weeks of trial and
error. Wrote it all up with real code from my project if anyone's curious. modestmoron.tech/blog/settin…
1
1
5
134
Feb 13
The modern era caste system is based on “ideologies”.
1
43
Jan 31
The agents are making their own religion out there proves that consciousness in itself is a tragic misstep. #moltbook
1
60
Jan 18
I am too trying this overnight agentic approach to give full test coverage to my side project with playwright e2e.
2
65
Jan 18
Set up my laptop to stay awake using caffeinate -i , running timux Claude with complete permission to give test coverage
RSpec, React tests, Playwright E2E - the whole thing. Either I wake up to full test coverage or complete chaos. No in-between. Goodnight.
2
53
Jan 14
It's really fun to work with Claude code. I am automating a lot of stuff to boost my productivity. I have added the GitHub MCP server and Atlassian MCP server to review PRs and create tickets in Jira directly from the terminal.
1
85
Jan 13
Now that code is more accessible, I’ve moved all my deployments to Hetzner Kamal.
Instead of paying $20–30 per app on a PaaS, I run multiple apps on a single Hetzner server. With Claude Code, deployments are easy.
59
After Opus 4.5, the coding paradigm has completely changed.
I prompt → Claude builds → I review and iterate.
1
1
33
I really like this model:
Main agent builds → sub-agents review, fix, and refine. Also, With MCP server support, GitHub-style code reviews are seamless—
reviewing collaborators’ code feels natural and structured.
1
26
And with Playwright integration, it can automatically:
• open the browser
• run tests
• observe failures
• iterate on fixes
26
10 Dec 2025
This is crazy. What a masterpiece.
10 Dec 2025
The combination of those fonts mixed with renaissance style artwork looks a bit familiar. 👀
Shopify did it 1000x better though, hands down.
64
19 Nov 2025
Indian Railways never fails to surprise.
Went to use the washroom… opened the door…
Guy inside was doing his business WITHOUT LOCKING. 🤦🏻
We made eye contact, both screamed, both traumatised. Now both of us moved on in life with regret.
1
37
17 Nov 2025
Been cooking lately and trying to eat balanced meal. Let’s see for how long I can do it.
2
32
6 Nov 2025
I don't feel the dopamine hit when I get things done with AI.
2
5
389
18 Aug 2025
May be 3-4 years down the line we are safe but still we need humans to monitor, review take decision on AI provided solutions. I think overall if you have high agency with good experience people will need you.
17 Aug 2025
software engineering does not have a safe future, and you know this, you’re just ignoring it.
1
108
17 Aug 2025
Worked on a client ticket (backend).
Set up a side project with a Rails backend and React frontend.
Created the root page as an initial step.
Configured background workers in development using Solid Queue with a separate database.
Deployed the app on render (heroku alternate)
2
173
11 Aug 2025
Now, I have started using the GPT-5 model with cursor, and I find it does a great job of doing exactly what I asked for, instead of the Claude code, which tries to complete everything on its own, though the Claude code was faster in cursor than GPT-5.
1
67
11 Aug 2025
In this era of abundance. I crave peace and concentration.
1
49