simple10

6 ms·

1.
by simple10 3mo ago
I use an observability tool with claude code [1] that shows me usage including prompt and session cost. Even though I use a max subscription, it's interesting to see what it would cost me if I was using API directly. My typical session
2.
by simple10 4mo ago
The Ryzen AI Max 395 128gb is super cool, but not fast for inference. Order of magnitude slower than dedicated GPU but at half the cost. You can run larger models on it but it's slow. Great for local async work. Not great for daily cha
3.
by simple10 5mo ago
The website has a broken github link. Repo is here https://github.com/DO-SAY-GO/ghostbox-releases but it's only a README.md with binary releases. No public source code?
4.
by simple10 5mo ago
I got kinda obsessed with observability a month ago and wired together a full stack for personal use. https://github.com/simple10/agent-super-spy - llm proxy + http MiTM proxy + LLMetry + other goodies https:/&#x
5.
by simple10 5mo ago
This is useful if you want to keep an eye on what claude's actually doing behind the scenes: https://github.com/simple10/agents-observe
6.
by simple10 5mo ago
They kinda buried the code deep in their docs: https://github.com/GoogleCloudPlatform/scion
7.
by simple10 5mo ago
Really cool. I've been building a mission control system (multi agent orchestration) that follows very similar patterns of spec driven development, steering, and task management. Having this baked into an IDE is a great idea. For obser
8.
by simple10 6mo ago
Yep. I finally realized what "green" accounts are for on HN. Recently created accounts.
9.
by simple10 6mo ago
Right on. Good luck! You might also want to play around with https://github.com/simple10/agent-super-spy if you want to see the raw prompts claude is sending. It was really helpful for me to see the system prompts and
10.
by simple10 6mo ago
Sub-agent trees are fully tracked by the dashboard. When an agent is spawned, it always has a parent agent id - claude is sending this in the hooks payload. When you mouse over an agent in the dashboard, it shows what agent spawned it. Ther
11.
by simple10 6mo ago
Sort of. It wasn't really noticeable until I did an intentional audit of performance, then noticed the speed improvements. Node has a 30-50ms cold start overhead. Then there's overhead in the hook script to read local config files
12.
by simple10 6mo ago
I hit a lot of limits on Pro plan. Upgraded to Max $200/mo plan and haven't hit limits for awhile. It's super important to check your plugins or use a proxy to inspect raw prompts. If you have a lot of skills and plugins inst
13.
by simple10 6mo ago
I'm not actually reading the jsonl files. Agents Observe just uses hooks and sends all hook data the server (running as a docker container by default). Basic flow: 1. Plugin registers hooks that call a dump pipe script that sends hook
14.
by simple10 6mo ago
This tool is useful if you want to see all the internal commands claude agents are making in real-time: https://github.com/simple10/agents-observe
15.
by simple10 6mo ago
Thanks! This was step one in my daily driver stack - better observability. I also bundled up a bunch of other observability services in https://github.com/simple10/agent-super-spy so I can see the raw prompts and heade
16.
Show HN: Real-time dashboard for Claude Code agent teams(github.com)
77 points by simple10 6mo ago|28 comments
17.
by simple10 6mo ago
Yes, this. They need as much lock-in as possible before IPO. Most likely less about cash flow and more about IPO story telling. We'll know for sure when they add full OpenClaw-like features to Claude Code like remote channels & hea
18.
by simple10 6mo ago
It's on my claw list to write a blog post. I just keep taking down my claws to make modifications. lol Here's the full (unedited) details including many of the claude code debugging sessions to dig into the logs to figure out what
19.
by simple10 6mo ago
NemoClaw is mostly a trojan horse of sorts to get corporate OpenClaw users quickly ported over to Nvidia's inference cloud. It's a neat piece of architecture - the OpenShell piece that does the security sandboxing. Gives a lot mor
20.
by simple10 6mo ago
Yeah, it's wild. I spent several weeks nearly full time on a deep dive of claw architecture & security. The short of it - OpenClaw sandboxes are useful for controlling what sub-agents can do, and what they have access to. But it&#x
21.
Show HN: LLM Observability Stack for Local Dev – Agent Super Apy(github.com)
2 points by simple10 6mo ago|0 comments
22.
by simple10 6mo ago
Klaus looks great! It's definitely looks like a step up from the one-click VPS deploys that are terribly insecure. I spent the past month hacking on openclaw to play nice in a docker container for my own VPS use. This project has a lot
23.
by simple10 7mo ago
Yeah, claude seemed genuinely shocked when it analyzed the logs. lol
24.
28M Tokens for One Simple OpenClaw Prompt(github.com)
3 points by simple10 7mo ago|3 comments
25.
by simple10 7mo ago
Just sharing the most surreal 3-hour claude assisted debugging session I've ever had. This was intended for my team notes but thought I'd share it. It's a treasure trove of how OpenClaw sandboxing works when things go wrong.
26.
by simple10 8mo ago
Also this one is trending on HN. https://skillregistry.io/ https://news.ycombinator.com/item?id=46692692
27.
by simple10 9mo ago
It's possibly related to mistaken identity. There's apparently some guy with a similar name that has some tie to Israel. Articles and social just seem to be running with it without any fact checking.
28.
by simple10 9mo ago
The article might be too hasty to report that he's Jewish, especially implying that was the motive by including it in the article title. Lots of chatter X/Twitter about it. Kinda crazy (scary?) how fast tragic events like this get
29.
by simple10 9mo ago
Here's the local Boston news reporting on it: https://www.youtube.com/watch?v=CmbmBNre5SQ
30.
by simple10 10mo ago
Maybe "Pingachu" if you want to go in the opposite naming directly. lol

More ›