Michael Born (@michaelborn_me) header
Michael Born (@michaelborn_me) profile picture

Software Engineer. I make search go brrr!

Upstate NY
Joined June 2013
Photo by Michael Born (@michaelborn_me)Photo by Michael Born (@michaelborn_me)Photo by Michael Born (@michaelborn_me)Photo by Michael Born (@michaelborn_me)Photo by Michael Born (@michaelborn_me)Photo by Michael Born (@michaelborn_me)Photo by Michael Born (@michaelborn_me)Photo by Michael Born (@michaelborn_me)Photo by Michael Born (@michaelborn_me)Photo by Michael Born (@michaelborn_me)
If you're wondering why the heck a smart guy like me (admittedly a stretch) is programming in Coldfusion as opposed to PHP or Node - here you are, you can wonder no longer. :) michaelborn.me/entry/why-cf
7
15
2
53
New Immich wallpaper app for the Samsung Tizen TV: * Pair via phone to handle the Immich connection details (no typing API key via TV remote) - Rotate through Favorites/album/All - Navigate slideshow with your tv remote - back/next/play/pause github.com/michaelborn/Immic…
1
125
Using Opencode Serve to build an immich app for my Samsung TV. - write instructions from desktop - check progress on phone I appear to be living in 2026.
1
1
60
Interesting- all 2 votes say they create zero new skills per week. I find myself creating at least 1 a week to improve Ai model usability on the legacy code uses i work with.
How many agent skills do you create a week? Just curious. #ai #skills
100%None
0%1
0%2
0%3+
2 votes • Final results
1
37
Michael Born retweeted
How many agent skills do you create a week? Just curious. #ai #skills
100%None
0%1
0%2
0%3+
2 votes • Final results
1
1
75
If you track code coverage, what files do you track? Do you track config files? Migrations? Docker compose files? I love TDD, but where do you strike a line between useful coverage metrics, and unhelpful metric chasing? #tdd
1
38
How many agent skills do you create a week? Just curious. #ai #skills
100%None
0%1
0%2
0%3+
2 votes • Final results
1
1
75
Batching MSSQL inserts to stay under the 2100-parameter limit imposed by the MSSQL driver. Am I too wordy with the AI instructions?
36
What if I could find out what this tool does without watching a full-length feature film while waiting for the page to load? @jonathan_wilke not meant to sound harsh. I want to learn more about Maible, but I can't wait that long for the "about" and "docs" links, etc. to load.
1
40
Folks, I did it. Am now working multiple Jira tickets at once using @herdrdev , #pi and @opencode . What a time to be alive.
1
37
This has been my question for quite a while now. What, exactly, are you building that takes more than a day to ship?
People who are running "agent swarms in the cloud" while they sleep, what's the use case? Even if I had a backlog of 100 issues I'm pretty sure a single Claude instance could rip through that in 30 mins let alone multiple, all night. Curious to know what people use this for.
1
38
Hot take: If creating an account auto-registers you for a newsletter, your newsletter is spam. #sorrynotsorry
1
28
When the entire internet is the input, it's nearly impossible to secure. The best protection against prompt injection is to limit the actions an AI agent can take. * Readonly DB access * File system jails * Containerized agents etc. dev.to/james_anderson_h/prom… #ai #security
3
1
2
90
Using Ghostty? Set light and dark themes so ghostty auto-switches theme to match your Ubuntu system theme. (Peeved you have to restart all ghostty windows for this to take effect, though!) github.com/michaelborn/today…
2
58
Painless is one language I cannot wait for AI to wrest out of my hands. #elasticsearch
36
Sometimes the most useful thing a leader can say is: I believe you. I don’t need the details. Tell me what we're changing. michaelheap.com/i-dont-want-…
23
Shout out to Chezmoi for making dotfile management so easy! chezmoi.io/
27
I just asked GPT-5.6 Luna and Claude Opus 5.5 to track down an issue in a client app. They BOTH discovered the issue, but while Luna got mired in unrelated tangents, Claude made the single change and responded with a simple report telling me What and Why. Thanks, Claude. #ai
1
62
For me, using Claude on more complex changes makes the difference between "DIY" and "AI, Why?".
15