Post

this janky always-on agent i vibed together from some of your wild ideas (forgetting, self-steering, personas, etc.) on top of pi durable is maybe the best agent experience i've had yet 🫪

i think next is something like amp's runner, where i or the main agent can easily install the harness to a new machine/VM and all the installations can collaborate with each other. maybe the runners only spawn subagents at the request of the main agent, wherever it happens to be? 🤔

i love telling one amp thread on one VM to go do something on another VM it only has access to through my amp account (no tailscale, no ssh) and it does by telling a new thread to do the thing. in my harness, that would be like instantiating another main agent, and i only want one main agent for now

0 reposts0 likes

i really want to do this with albedo and atproto oauth but the issue is with spaces is that the pds owner knows everything thats going on

expect everyone to self-host! /hj you could generate keypairs and store the pub part in your PDS like tangled does for knot SSH access, but that seems pain for this. if you like tailscale, you could use tsnet to put every instance on your tailnet, then security of the messaging system between…

nodes doesn't matter so much because ts handles auth, its ACLs constrain traffic flow and those constraints can apply immediately on adding a node by adding it with the right ACL tag(s), traffic is encrypted over the interwebz, you can share an instance with oomfs and give them different…

controls or access or whatever based on identity headers i planned to do some of this until deciding on pi durable as a base and foregoing tsnet