Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
davidcann
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
1.
▲
by
davidcann
3mo ago
You can block file access on macOS by running codex in a sandbox with my app: https://multitui.com
2.
▲
by
davidcann
3mo ago
It works for me, but only if I hold the phone up in front of my face, basically perpendicular to the road.
3.
▲
by
davidcann
3mo ago
Thanks for the report. I messed up the CDN settings. It looks fixed now.
4.
▲
by
davidcann
3mo ago
I built this with 94% written by coding agents: https://buildermark.dev/ The complete log of all prompts and commits is here: https://demo.buildermark.dev/projects/u020uhEFtuWwPei6z6nbN
5.
▲
by
davidcann
3mo ago
Photoshop alternatives already exist and are cheap or free, like Affinity. Why create a Final Cut Pro alternative when it’s so cheap from Apple? Many existing apps can be mostly cloned by a small team over 6 months or a year, but the challe
6.
▲
by
davidcann
4mo ago
You can also sandbox TUIs, with full control of the sandbox parameters. To do this on macOS, you can use: https://multitui.com
7.
▲
by
davidcann
4mo ago
Buildermark calculates how much of your code is written by agents. Open source, local, and cross-platform (written in Go). https://buildermark.dev
8.
▲
by
davidcann
5mo ago
> 70% of committed code originating from AI. How are they calculating that? They could be using my tool, Buildermark, but I do t think they are: https://buildermark.dev
9.
▲
by
davidcann
5mo ago
Thanks for the feedback. I agree that it’s changing very fast, which is why my thesis is that this tooling will be needed to help everyone on the team keep up.
10.
▲
by
davidcann
5mo ago
Is your team measuring how much of your code is being written with claude and comparing amongst the team, like what works best in your codebase? How are you learning from each other? I’m making a team version of my buildermark.dev open sour
11.
▲
Calculate how much of your code was written by AI
(github.com)
1 points
by
davidcann
5mo ago
|
0 comments
12.
▲
by
davidcann
5mo ago
To help you decide if you should keep your Claude subscription, you can see how much of your code is written by Claude Code with my project (open source, local): https://github.com/gelatinousdevelopment/buildermark
13.
▲
See 364 prompts that wrote 94% of Buildermark
(demo.buildermark.dev)
2 points
by
davidcann
5mo ago
|
0 comments
14.
▲
by
davidcann
5mo ago
I’m working on Buildermark (open source, local) that calculates how much of your code is written with coding agents. It scans your claude and codex history to find edits and matches those to git commits (even if the code was auto-formatted)
15.
▲
Show HN: Buildermark – See how much code is by your agents (open source, local)
(buildermark.dev)
2 points
by
davidcann
5mo ago
|
0 comments
16.
▲
by
davidcann
6mo ago
Native macOS sandbox terminal: - UI for sandbox-exec to protect filesystem - Network sandbox per domain - Secrets filter via gitleaks - Vertical tabs option It's highly customizable. You generate native macOS app wrappers for each term
17.
▲
by
davidcann
6mo ago
I made a native macOS app with a GUI for sandbox-exec, plus a network sandbox with per-domain filtering and secrets detection: https://multitui.com/
18.
▲
by
davidcann
7mo ago
No, I run a separate URL detection to make links clickable. However, SwiftTerm just added link detection a few days ago and I haven’t had time to look into theirs yet.
19.
▲
by
davidcann
7mo ago
My app is a macOS terminal wrapper with nice GUI for sandbox-exec and network sandbox. I just added a vertical tabs option too. https://multitui.com
20.
▲
by
davidcann
7mo ago
Yeah, that’s understandable. Many open source macOS-only apps seem to get abandoned, so I’m trying to build something sustainable. It uses only 3 dependencies that are very well known and widely used, so supply chain risk is minimal. That l
21.
▲
by
davidcann
7mo ago
SwiftTerm, for not. I may switch to ghostty when their library is ready for embedding.
22.
▲
by
davidcann
7mo ago
Yeah, they all do sometimes, but the agent decides what to allow and they can choose to not use it. This gives the user full control of the sandbox and you can run the agent in yolo mode.
23.
▲
by
davidcann
7mo ago
I made a UI for this to run terminal apps, like claude and codex: https://multitui.com
24.
▲
by
davidcann
8mo ago
The same concept is possible on Linux, but I don't think anyone has created a nice UI for it yet. There was a post yesterday about doing it on the command line in linux: https://news.ycombinator.com/item?id=46874139 On
25.
▲
Show HN: Multitui – sandbox claude/codex/gemini on macOS without containers
(multitui.com)
1 points
by
davidcann
8mo ago
|
3 comments
26.
▲
by
davidcann
8mo ago
I’m seeing log delays too the past couple of days… something like 14 hours, in my case.
27.
▲
by
davidcann
8mo ago
My app does this on macOS! https://multitui.com
28.
▲
by
davidcann
8mo ago
Thanks! For network sandboxing, I was thinking something like what Little Snitch can do, but more customized... maybe block POST requests and long GET request strings or anything that looks like too much code or secrets.
29.
▲
by
davidcann
8mo ago
Shameless plug, but you can sandbox codex cli without a container using my macOS app: https://multitui.com
30.
▲
Show HN: Multitui – sandbox claude, codex, or any TUI with custom macOS apps
(multitui.com)
2 points
by
davidcann
8mo ago
|
2 comments
More ›