But that's how taste works! It's personal
3,393 karma · joined February 13, 2020
writing: yakko.dev gh: https://github.com/yakkomajuri
But that's how taste works! It's personal
The rollercoaster on the second one was the breakthrough that then eventually led to what I have today
As an example, I recently built my company's website and I'm really happy with the result. I wanted it to have some hints of Rollercoaster Tycoon, and with a lot of experimentation, direction, tinkering, and examples I got to something that I really liked. Two years ago my company website would have been some generic Bootstrap thing until I hired a designer.
There's a lot to improve and certainly some aspects of it feel like AI still but it's generally gotten a lot of positive feedback and above all I really like it. Considering writing a blog post about my creative process here because I'm a dev with no design background at all.
Sharing a screenshot as to not promote: https://imgur.com/a/xqKi6X6 (in the website everything in the park moves btw)
Libraries are a place of possibilities and fun, and it makes people want to be there. You can imagine the long-term positive impact this has.
I added plugins this year which made it really powerful and allows me to keep the core small.
Basically, build anything using your agent of choice (Claude Code, Codex, etc) and get it live in your company's infra without having to provision anything yourself / loop in a DevOps person.
App goes behind company auth (e.g. SSO), and everything's handled for you: custom domains, governed storage APIs, observability, analytics, role-based access control, etc.
Private Alpha atm.
P.S. Sample size is currently small but I've also gathered a fair amount of qualitative data to indicate both of these things. Of course measuring productivity is also far from a science.
Tiny bit of feedback on the demo video: the beginning makes me dizzy and I'm not sure what's going on (was watching on my phone).
Although of course returning is a weird term in the sense that the attackers will almost certainly keep the data as well.
Annoyingly the web is becoming a bit more annoying to browse as a DuckDuckGo (mobile) and Brave (desktop) user. With a VPN on top it gets even worse.
I don't even remember how we used to split bills back in the day anymore.
However, I've been working remotely for 7 years now and recently became a solo founder, and I realized I'm having a fair amount of social anxiety. At the previous two companies, I was working remotely but still had people online to chat to, and would meet in person once in a while. Now as a solo founder I've just been working from home and I noticed that when I was leaving the house to buy groceries or work out that was my "break time" and I somehow just wanted to be more alone so I always had my headphones on.
That meant that I became someone who's running away from social interaction the more I actually needed it. And that when placed in a social situation I'm suddenly anxious whereas before this all came very naturally to me (I've also spoken in public very often etc).
How I'm coping:
- Got a WeWork membership
- Leaving the house without headphones
- Striking up conversation with uber drivers, cashiers, etc
- Making an effort to go to events (even flying somewhere at my own expense to speak at a small event for the first time in years)
Plus the execution is also part of the art.
I saw you're using Gemini for the sentiment rating (which I guess you picked because it's not often mentioned and thus "neutral"? lol)
But would be interesting to get more details overall
So the difference here is the granular permissioning layer. In practice it means AgentPort is basically an open source Composio with granular permissions today. But whereas their focus is on just connecting you to everything my focus here is security. And that leads to pretty different products down the line in my opinion.
As for connecting your own tools, I have a feature basically ready to let you create your own integrations. Should launch soon!
So today there are two modes:
1. The agent halts and waits for you to approve
2. The agent can fire the tool calls and keep working on other things and only at the end ask for the approvals it needs
It's not perfect, but it does work well in a lot of cases. Retrying is pretty simple because you can just tell the agent to re-run the same tool call and approve it then. As of right now approvals expire within 15min but I'm actually putting a PR up today to make this configurable.
Adding a human in the loop technically does limit how much can be done fully autonomously (e.g. via scheduled agents) but: a) I think the tradeoff is worth it b) it gives me peace of mind to let agents run unsupervised on a schedule for tasks that shouldn't require approval because I know they won't do something they weren't instructed to do when left unsupervised.
Nevertheless, it's a tricky design problem. I prefer the current approach to having to log into AgentPort and approve things from a "pending approvals list" there, but what do you think? What would make most sense to you?
One of the principles I believe you should follow is: if there's enough access for an action to be taken, then you must assume that action can be taken at any point.
Basically, if it has access to delere prod data, you should assume it might do it and plan accordingly.
I also believe the actions of your agent are entirely your responsibility.
As part of my digging into securing these systems I've baked some of these principles into AgentPort, a gateway for connecting agents to third-party services with granular permissions.
If anyone's interested in this space:
EDIT: it does seem to have a deterministic layer too and I think that's great
I think you're spot on with the fact that it's so far it's been either all or nothing. You either give an agent a lot of access and it's really powerful but proportionally dangerous or you lock it down so much that it's no longer useful.
I like a lot of the ideas you show here, but I also worry that LLM-as-a-judge is fundamentally a probabilistic guardrail that is inherently limited. How do you see this? It feels dangerous to rely on a security system that's not based on hard limitations but rather probabilities?
https://github.com/yakkomajuri/seams
Run `seams .` in any dir and get a rich markdown editor with image uploads, block editing, tables, etc etc
Congrats on launching!
It was really easy to setup and I've been getting some value out of it but hasn't been the craziest thing in the world. I'm using it for:
- Unstructured note-taking: I suck at notes and todos and used to have a WhatsApp chat with myself (this is really common in Brazil) where I dump stuff. Now I dump into Hermes and it sorts whatever I put in there into one of various lists like to-do, to-read, to-try, to-buy, and so on.
- Briefings on a cron: I get reminded of my todos every morning and at the end of the day so I can cross stuff off. Later in the day I get reminded of my to-read list. I also get a summary of what went on from my coding orchestrator.
- Some coding: I built my own remote orchestrator and have been using Hermes to manage tasks, review code, and trigger tasks when on the move. Hermes has been a nice interface to allow me to use the orchestrator on my phone.
Haven't connected email or anything else yet. I feel like the security story here is lacking.
Overall it's been interesting but not mind-blowing. Plus setting up was easy but it's a bit buggy at times, messing up where files were and not being able to configure itself according to its own docs.
EDIT: Ah yes and voice notes via WhatsApp out-of-the-box is really nice
Congrats, and I'm rooting for this to work.
Nice post.
(P.S. I'm also Finnish)