gotgenes/pi-permission-system and the ai-guard authorizer can do a lot here, it feels very similar to cc auto mode.
I built a policy authorizer on top of pi-permission-system that looks kinda like nftables that models commands/subcommands/options as read/write and enfores policies depending on if the sub agent is read only or write, which really cuts down on the stupid decisions and nicely complements AI cmd review. Pi shines for making things like that possible.
I have a time of use electric plan and adjust load (grow lights, dehumidifier, AC, battery backups) at certain times to save money. Smart home can be pure tinkering but it can control some pretty practical things.
You do you though, mechanical outlet timers work perfectly well.
There are much better ways of device enrollment; at a minimum they could require device activation that doesn't blindly use a token with no further checks.
I'm concerned about publicly accessible devices containing secrets also. These are not physically secure places to store keys, they are mounted on street lights where anyone with a ladder can get the key material out of the device.
* Privacy, civic trust, society if you get a chance!
This is the end product of tech leadership taking fat rips of disruption cocaine for the last 15 years. Flock Safety got VC money so that they could build a panopticon. There is nothing surprising about the fact that they did a hack job with terrible security; the fact that their service names are various types of alcohol is beyond parody.
Oh well, at least Flock Safety's IPO will be a critical cash infusion in the pursuit of building the torture nexus so that's cool.
In this metaphor the police are corrupt, running a protection racket, and might get drunk and shoot you while your dealer is in the mafia. The mafia link might be a problem at some point but the person with the badge and the gun is a problem right now.
Manufacturers could add modems to bypass the need for a wifi or cable connection, like they do in cars. I would expect by now that's a valid question to ask if I were to buy a new TV.
In practice this requires duplicating the upstream build and release pipelines, so now you need yet another agent running in the background (+ periodically going off the rails + eating tokens) which incurs more up front cost and cognitive overhead.
In my experience "throw another agent at it/fork it" consistently creates unsupervised trainwrecks.
Fully agree with this. I did that for pi for a while, maintained it and brought merges from upstream while having my own patches on top but then went on holidays and there was that refactor where the agent pipeline failed... now I'm stuck on the version from april/may (works great but I can't use extensions; and I'm too lazy to debug/fix while everything works great).
off: I'm working on my spare time on a code mode lisp alternative (great opportunity to learn lisp) and might switch to it fully as long as I built some simple evals (I'm concerned about token usage, which is why i forked pi the first time)
reply