HackerTrans
TopNewTrendsCommentsPastAskShowJobs

aymenfurter

no profile record

Submissions

[untitled]

1 points·by aymenfurter·2 miesiące temu·0 comments

Show HN: Strava for AI coding – analytics on your Copilot/Claude/Codex usage

github.com
8 points·by aymenfurter·2 miesiące temu·1 comments

[untitled]

1 points·by aymenfurter·5 miesięcy temu·0 comments

[untitled]

1 points·by aymenfurter·5 miesięcy temu·0 comments

[untitled]

1 points·by aymenfurter·6 miesięcy temu·0 comments

Ralph for GitHub Copilot

github.com
2 points·by aymenfurter·6 miesięcy temu·0 comments

Show HN: Ai3 – An experimental agentic tiling window manager (i3 fork)

github.com
4 points·by aymenfurter·7 miesięcy temu·1 comments

Environment Engineering: The Missing Layer in AI Agent Design

aymenfurter.ch
16 points·by aymenfurter·9 miesięcy temu·0 comments

Voice Live API: AI Sales Coach (Azure Sample)

github.com
1 points·by aymenfurter·10 miesięcy temu·0 comments

comments

aymenfurter
·2 miesiące temu·discuss
Chess banned engines from competition. CTFs can't really do that because you need internet access and tooling to play.
aymenfurter
·2 miesiące temu·discuss
PG is proof that the best things on the internet are still built by people who just care about the mission.
aymenfurter
·7 miesięcy temu·discuss
I built ai3 to control my window manager via chat by exposing the i3 IPC protocol as a toolset for LLM agents. It includes a Next Action prediction feature similar to the tab-complete in AI code editors that anticipates your next action based on current activity. You can also trigger AI based layout optimization and control your desktop through chat. This is an experimental research project so please use it for testing only and be mindful of your data privacy.
aymenfurter
·7 miesięcy temu·discuss
Big jump from gpt-image, but I would love to see the same kind of step change we have seen in recent language model releases.
aymenfurter
·9 miesięcy temu·discuss
Async agents are great. They let you trigger work with almost no effort, and if the results miss the mark you can just discard them. They also run in the background, making it feel like delegating tasks across teammates who quietly get things done in parallel.