Live data from Hacker News

Ask HN: What are you working on? (August 2026)

news.ycombinator.com

451–460 of 1001 posts

Re: Ask HN: What are you working on? (August 2026)

#455
I am building this:

https://safetyevidence.org

It is the most comprehensive collection of benchmarks of the safety, ethics, and values of LLMs. I also aggregate sparse safety benchmark scores into a global ranking by optimizing the global ranking to maximize correlation with individual benchmarks.

Re: Ask HN: What are you working on? (August 2026)

#456
Since January I've been working on a storytelling system that fuses what I think is the best of interactive fiction / MUDs with the strengths of LLMs. The world model is entirely deterministic, but the LLM interprets the player and narrates the game and NPCs.

Also hacking on something similar to YC's QM but designed for small non-profits to ease as much back office work as possible.

Re: Ask HN: What are you working on? (August 2026)

#458
I’m building cloud managed self improving agents at https://sanbox.cloud.

Each instance gets its own persistent file system, root SSH access and a web terminal. The goal is to let an agent build skills with experience rather than start from scratch on every task.

I started with Hermes as the first harness and now exploring Prime Agent. You can use it for a personal assistant, use it to build LLM wiki, connect it as part of an agent team, or run any coding agent on the cloud with dangerously skip permissions.

I’m also working towards a more general-purpose agent setup that can connect to channels such as email, WhatsApp, Telegram, Slack, Buzz.

Re: Ask HN: What are you working on? (August 2026)

#460

I'm building an uptime monitoring service, in Elixir, called https://larm.dev . It's a tough field because for some reason people vibe coding apps think this is very easy to do, and there are new apps coming out pretty much every day. I'm still figuring out how to effectively stand out in the noise. I really enjoy working on the app though. The backend is Elixir, with a Phoenix LiveView dashboard. I've got a cluster…

Elixir is perfect for stuff like this, very cool project!
Post reply on HN