I wanted to manage long horizon agentic workstreams via voice, then put my phone down, and have a harness manage completion - extending into full computer use.
I was trying to build a personal Jarvis, so I benchmarked AutoBot to see how close I could get:
- OSWorld: 32.41% (moved Sol Max from 4th to 1st, beating Opus 5)
- AssistantBench: 50.70%
Hermes and OpenClaw, but without needing a weekend and VMs. And with the ability to manage deep personalization AND keep strict privacy rules, storing data on encrypted disk.
AutoBot is a passion project that grew out of trying to make this all work for myself.
I’m sharing it here because I suspect other here have the same frustration. And I’m curious what else everyone is doing for this.
It’s an MIT-licensed harness that lives within a project.
Native voice lets me discuss tasks, check progress, and steer work; a local ledger tracks unfinished outputs and the evidence needed to call them done. Memory drives more autonomy over time, and defrags and locks in learning nightly while a heartbeat system persists execution.
I’m not selling anything. If you’re building something similar for yourself, I’d love to compare notes.
Thank you for sharing! It's funny, yesterday I just built an iPhone Shortcut that allows me to dispatch async claude code tasks, i.e. fixing a bug just by saying "dispatch go" on my phone. Really cool to see other people working on stuff like this!
Yeah, that is a perfect use case for commuting, going for a walk in between meetings, etc. Same concept here.
But then the next annoying thing is the endless approval and safety nags. So there is a somewhat hacky skill I posted on my GH that is designed to trick the models into approving each other (use cautiously, it can go sideways fast in full computer use mode).
Ha, I know. The way I end up solving it is to give full permissions to the claude/codex session and just try to lockdown any external actor from reaching my systems, but there's still a risk of course. The tools I work on are internal enough that I don't lose too much sleep over it, but prompt injections still scare me.
I was trying to build a personal Jarvis, so I benchmarked AutoBot to see how close I could get:
- OSWorld: 32.41% (moved Sol Max from 4th to 1st, beating Opus 5)
- AssistantBench: 50.70%
Hermes and OpenClaw, but without needing a weekend and VMs. And with the ability to manage deep personalization AND keep strict privacy rules, storing data on encrypted disk.
AutoBot is a passion project that grew out of trying to make this all work for myself.
I’m sharing it here because I suspect other here have the same frustration. And I’m curious what else everyone is doing for this.
It’s an MIT-licensed harness that lives within a project.
Native voice lets me discuss tasks, check progress, and steer work; a local ledger tracks unfinished outputs and the evidence needed to call them done. Memory drives more autonomy over time, and defrags and locks in learning nightly while a heartbeat system persists execution.
I’m not selling anything. If you’re building something similar for yourself, I’d love to compare notes.
GitHub: https://github.com/demeyer1/Autobot
Thank you for sharing! It's funny, yesterday I just built an iPhone Shortcut that allows me to dispatch async claude code tasks, i.e. fixing a bug just by saying "dispatch go" on my phone. Really cool to see other people working on stuff like this!
Yeah, that is a perfect use case for commuting, going for a walk in between meetings, etc. Same concept here.
But then the next annoying thing is the endless approval and safety nags. So there is a somewhat hacky skill I posted on my GH that is designed to trick the models into approving each other (use cautiously, it can go sideways fast in full computer use mode).
Ha, I know. The way I end up solving it is to give full permissions to the claude/codex session and just try to lockdown any external actor from reaching my systems, but there's still a risk of course. The tools I work on are internal enough that I don't lose too much sleep over it, but prompt injections still scare me.