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.

I did the same thing. I use the Pi coding agent and they support extensions. I asked GPT 5.5 to create a Pi extension that has the magic of Hermes and for me that was the long term memory and the cron/scheduling. And that way, I can use Pi as is and turn on these using a slash command.

I've done similar, even to the point of a discord gateway integration that is very similar to the one Hermes ships with.

I still use Hermes because it's not a problem I care about solving right now and am spending my time and tokens on more important problems. I don't plan to always use it, and if anything it gives me ideas for what the ideal shape of what I want would be.

Also to add, having a unified space to chat with an agent that has a local system it can work with is quite liberating because I use it to drive my agent orchestration system. So I'm no longer bound to my computer or any one providers harness and whatever mobile solution they provide if any. It's been quite freeing and I don't use Hermes that differently than I would chatgpt, except it know exactly how I like to do my work and that means I have to remind it less.

[flagged]