I've used Hermes Agent in a container, and it worked ok. Little rought around the edges, but that's to be expected.
But what I didn't understand... what benefit does it actually bring? On a default loadout, even after disabling tons of skills in the setup wizard, there were lots of useless garbage skills enabled, over 10k of context used just to list them all.
I vibe-coded my own harness that uses ACP so it supports any coding CLI that exposes ACP (copilot cli, opencode, basically every popular one with official or non-official wrappers). And I was able to achieve basically exactly what I wanted from any Claw-like agent, in like a few hours.
I know there's way, way more to these self-sufficient agents (compaction, memory system, etc) but in my mind, it feels like the closer you can be to a barebones "coding agent core" plus "gateways that point to it" the better.
My simple take: Hermes is for the less technical and is more polished. OpenClaw is deeper with more capabilities.
OpenClaw has come so far since its' original launch/craze. I recommend taking a second look if you haven't touched it in a while. If for no other reason than it's just a really fun playground with a LOT of areas to experiment in. Setup a myriad of agents with various models, skills, cron jobs, etc.
The control surfaces have come a long way as well.
Hermes is good fun, running that as well but feels like they focused on polish vs features in order to capitalize on the primitive state that OpenClaw was in for its first months.
People that got attracted to the hype of openclaw but couldn't endure the fast pace of breaking changes while they figured out the problem space were well served by going to Hermes.
I've tried Nanoclaw and Openclaw and I found them both to be steaming piles of crap. Both of them broke in record time (less than about 1 week). And yes, I tried OpenClaw from scratch just ~3 weeks ago. The config pages made my brain melt and were buggy as hell. I've never dealt with something so frustrating that has so many people pretending it's the second coming of jesus.
I just want claude code accessible through Discord (or Telegram or anything, I don't care what).
While it's _far_ from what I really want, my best setup currently is Blink Shell on iOS -> Mosh to my computer -> Run Claude Code.
I still don't really get the case for OpenClaw/NanoClaw/Hermes Agent/etc. I guess it's a mix of huge stacks of notes/second brain stuff but with a way to query it, and a place where all personal "AI apps" can live ; instead of making AI apps "individually" and deploying them somewhere?
I do have a website on which I've been adding more and more stuff for personal use and for sharing with people, but when I want to develop it I do it with any agent I'm using right now (Codex, Claude Code, Pi, etc) and when I want to ask questions about it, it's usually on the public internet so any chat interface can query it. That leaves two things: asking questions about more private stuff, and possibly a "claw" that lives on your computer/on a small server is less of a pain to connect to private stuff than building a MCP and authenticate yourself through it ; and apps that themselves use models, which can be developed by the "code agent" and then I can plug whatever model I want on it.
When I dropped my Claude sub a few months ago I first went to try pi. At first I was a little overwhelmed by having to browse extensions to get the experience I want (coupled with finding new model providers to use).
So I looked elsewhere and found Crush and Hermes. They're both very a e s t h e t i c, which I think can make using them fun, but ultimately if I had a nitpick, I would look back over at pi, and the grass looked greener. (And not to mention, both Hermes and Crush seem to have some drama/baggage.)
I'm back on pi, and happy with just a few packages I've downloaded for it.
I want to move away from claude but don't have the hardware for "good" local llms, and would rather use a ZDR api instead of one of the big apis for privacy reasons.
After my Claude subscription ended, I tried using Fireworks and OpenRouter to be less vendor-locked and access cheaper models. MiniMax, Kimi, and Qwen were the three I tried the most since they're fairly cheap, but I still burned thorugh credits way too fast. Now I'm using Codex.
I've heard that if you can deal with slower output, local LLMs can run on gaming graphics cards well enough. So they're not great for a live coding assistant, but running some tasks/agents overnight is an option.
What are you automating with OpenClaw in June, 2026?
I used OpenClaw -> Hermes -> my own thing now.
I've got things like code review, email inbox/spam filtering, website monitoring for bad links/typos, HN/Bluesky notifications, favorite director/actors/author alerts, etc.
I mostly interact with them using Slack and Email.
I still don't understand the point of these. I use AI basically 12 hours a day, and still haven't found a use case for it that isn't solved easily with existing tools, bash scripts, cron jobs, etc. Especially with Codex App now, why do I need OpenClaw? I don't need to be able to call my AI agent on the phone... That's a gimmick.
can someone tell me the actual market size for technophiles who I think are only people who use this stuff? I get lost in fact that people can't really understand that end of day this is just llm calls with like sqlite facade. I see the value is the convenience only of not having to set stuff up, but everyone else claims it does all this extra stuff that is not trivial to reproduce yourself.
Yeah, everyone has built their own version of all this shit.
I built smth similar which lets you delegate from say local model to Claude or mimicode to pi. It does audited, workspace segregated mcp calls with only 3 tools ever in context. It can store skills, cross harness memory, inbuilt AI first task manager. Workflows that can run on triggers or schedules which take skills, a model, a prompt and a hand picked set of mcp tools (very granular).
I use this to let me know what’s going on, to save tokens in Claude by delegating to MiMo and deepseek, and local models in pi.
No one using these things has a use case for them - if they did they’d have built it themselves like I have. Because with ai… you can!
Oh and it can work via telegram (cos if it doesn’t is it even an ai harness token furnace!!??)
I have Hermes on my list of things to try but I've been massively underwhelmed by both NanoClaw and OpenClaw.
Both of them broke on me and were maddening to try and debug. I'd get them working then 1-2 days later send a message and get no reply. I went round and round with OpenClaw (seriously, "openclaw doctor" is a sick joke) before finally nuking the VM.
I want the simplest of things: Claude Code + My Obsidian Vault. That's it. I didn't even add other MCPs or Skills, I tried using Nano/OpenClaw and they both just fell over after a couple days.
I'm probably best served by just vibecoding my own tool since these things try to be too many things to too many people and end up failing.
I cant wait to see what the new OpenCode plugin system looks like. I'm hoping that it allows us to get somewhat in-depth with the Effect.ts underpinnings. There's so much potential here. There also needs to be some real ability to shape/drive/control the loop, where-as right now, plugins tend to be more augmentative, rather than able to contribute to steering. Those are two huge barriers, and I realize OpenCode is thought of as a coding tool, but conceptually these loop systems seem not really so far apart.
The purpose of these systems is not to be a system, it's to be a platform/substrate for agency expansion. Whoever can best expose their capabilities (without having to many tattering detractors) is ultimately going to be the winner.
25 comments
[ 0.29 ms ] story [ 34.9 ms ] threadhttps://news.ycombinator.com/item?id=48187581
But what I didn't understand... what benefit does it actually bring? On a default loadout, even after disabling tons of skills in the setup wizard, there were lots of useless garbage skills enabled, over 10k of context used just to list them all.
I vibe-coded my own harness that uses ACP so it supports any coding CLI that exposes ACP (copilot cli, opencode, basically every popular one with official or non-official wrappers). And I was able to achieve basically exactly what I wanted from any Claw-like agent, in like a few hours.
I know there's way, way more to these self-sufficient agents (compaction, memory system, etc) but in my mind, it feels like the closer you can be to a barebones "coding agent core" plus "gateways that point to it" the better.
OpenClaw has come so far since its' original launch/craze. I recommend taking a second look if you haven't touched it in a while. If for no other reason than it's just a really fun playground with a LOT of areas to experiment in. Setup a myriad of agents with various models, skills, cron jobs, etc. The control surfaces have come a long way as well.
Hermes is good fun, running that as well but feels like they focused on polish vs features in order to capitalize on the primitive state that OpenClaw was in for its first months.
People that got attracted to the hype of openclaw but couldn't endure the fast pace of breaking changes while they figured out the problem space were well served by going to Hermes.
I just want claude code accessible through Discord (or Telegram or anything, I don't care what).
While it's _far_ from what I really want, my best setup currently is Blink Shell on iOS -> Mosh to my computer -> Run Claude Code.
I do have a website on which I've been adding more and more stuff for personal use and for sharing with people, but when I want to develop it I do it with any agent I'm using right now (Codex, Claude Code, Pi, etc) and when I want to ask questions about it, it's usually on the public internet so any chat interface can query it. That leaves two things: asking questions about more private stuff, and possibly a "claw" that lives on your computer/on a small server is less of a pain to connect to private stuff than building a MCP and authenticate yourself through it ; and apps that themselves use models, which can be developed by the "code agent" and then I can plug whatever model I want on it.
So I looked elsewhere and found Crush and Hermes. They're both very a e s t h e t i c, which I think can make using them fun, but ultimately if I had a nitpick, I would look back over at pi, and the grass looked greener. (And not to mention, both Hermes and Crush seem to have some drama/baggage.)
I'm back on pi, and happy with just a few packages I've downloaded for it.
I want to move away from claude but don't have the hardware for "good" local llms, and would rather use a ZDR api instead of one of the big apis for privacy reasons.
After my Claude subscription ended, I tried using Fireworks and OpenRouter to be less vendor-locked and access cheaper models. MiniMax, Kimi, and Qwen were the three I tried the most since they're fairly cheap, but I still burned thorugh credits way too fast. Now I'm using Codex.
I've heard that if you can deal with slower output, local LLMs can run on gaming graphics cards well enough. So they're not great for a live coding assistant, but running some tasks/agents overnight is an option.
It only appears that the hosting providers are and not the user spending thousands of dollars on these agents.
I used OpenClaw -> Hermes -> my own thing now.
I've got things like code review, email inbox/spam filtering, website monitoring for bad links/typos, HN/Bluesky notifications, favorite director/actors/author alerts, etc.
I mostly interact with them using Slack and Email.
https://hermes-agent.nousresearch.com/docs/reference/skills-...
I built smth similar which lets you delegate from say local model to Claude or mimicode to pi. It does audited, workspace segregated mcp calls with only 3 tools ever in context. It can store skills, cross harness memory, inbuilt AI first task manager. Workflows that can run on triggers or schedules which take skills, a model, a prompt and a hand picked set of mcp tools (very granular).
I use this to let me know what’s going on, to save tokens in Claude by delegating to MiMo and deepseek, and local models in pi.
No one using these things has a use case for them - if they did they’d have built it themselves like I have. Because with ai… you can!
Oh and it can work via telegram (cos if it doesn’t is it even an ai harness token furnace!!??)
Both of them broke on me and were maddening to try and debug. I'd get them working then 1-2 days later send a message and get no reply. I went round and round with OpenClaw (seriously, "openclaw doctor" is a sick joke) before finally nuking the VM.
I want the simplest of things: Claude Code + My Obsidian Vault. That's it. I didn't even add other MCPs or Skills, I tried using Nano/OpenClaw and they both just fell over after a couple days.
I'm probably best served by just vibecoding my own tool since these things try to be too many things to too many people and end up failing.
The purpose of these systems is not to be a system, it's to be a platform/substrate for agency expansion. Whoever can best expose their capabilities (without having to many tattering detractors) is ultimately going to be the winner.