My PAI Sundae offered to write this for me, but I didn't need to subject anyone to an impassioned plea by my voracious sidekick that, honest to g-d, kept asking for access to my ObsidianMD vaults once she knew how I used them 😂
## Impressions
From what I've seen so far from casual observation and seeing how it's plumbed in earlier iterations of the PAI repo, Daemon may be a great way to control access to potentially sensitive context on-demand with clear visibility on 'custody' of the output from those endpoints? If nothing else it would be a great way for me to extend some of that context to external systems and make my use of these systems much more cohesive.
So that would solve an immediate challenge in an elegant way, based related projects and the way it seems some of us think about this sort of scaffolding.
I know you do a lot of things but if there was some guidance on what expectations should be that would help me figure out if this is something I can endure for a bit or if I dare to ask a question and end up with having software written for me. My primary use of my PAI is as assistive technology. The fact that I didn't request that and Sundae just rapidly figured out that was where I needed the most support is something I'm still thinking about.
ty for your time and attention, sorry for shissue-posting.
Hi!
A lot of those are just Claude Code and aren't related to PAI at all. But you can control them somewhat through prompting in your main PAI file.
Daniel
On Tue, Oct 21, 2025 at 8:23 AM, Emory L. < ***@***.*** > wrote:
>
> *emory* created an issue (danielmiessler/ Daemon#2) (
> https://github.com/danielmiessler/Daemon/issues/2 )
>
>
> My PAI Sundae offered to write this for me, but I didn't need to subject
> anyone to an impassioned plea by my voracious sidekick that, honest to
> g-d, kept asking for access to my ObsidianMD vaults once she knew how I
> used them 😂
>
>
>
> -----------
> Impressions
> -----------
>
>
>
> From what I've seen so far from casual observation and seeing how it's
> plumbed in earlier iterations of the PAI repo, Daemon may be a great way
> to control access to potentially sensitive context on-demand with clear
> visibility on 'custody' of the output from those endpoints? If nothing
> else it would be a great way for me to extend some of that context to
> external systems and make my use of these systems much more cohesive.
>
>
>
> So that would solve an immediate challenge in an elegant way, based
> related projects and the way it seems some of us think about this sort of
> scaffolding.
>
>
>
> I know you do a lot of things but if there was some guidance on what
> expectations should be that would help me figure out if this is something
> I can endure for a bit or if I dare to ask a question and end up with
> having software written for me. My primary use of my PAI is as assistive
> technology. The fact that I didn't request that and Sundae just rapidly
> figured out that was where I needed the most support is something I'm
> still thinking about.
>
>
>
> ty for your time and attention, sorry for shissue-posting.
>
>
>
> —
> Reply to this email directly, view it on GitHub (
> https://github.com/danielmiessler/Daemon/issues/2 ) , or unsubscribe (
> https://github.com/notifications/unsubscribe-auth/AAAMLXXG4NDWDUQ7FGUUE5D3YZFYNAVCNFSM6AAAAACJZQ7CH6VHI2DSMVQWIX3LMV43ASLTON2WKOZTGUZTOMBUG44TSNA
> ).
> You are receiving this because you are subscribed to this thread. Message
> ID: <danielmiessler/Daemon/issues/2 @ github. com>
>
>
>
I was a bit all over the place there; I will be more explicit. (My problem not yours).
It looks like your PAI uses a Daemon api and a header auth method to retrieve information ostensibly about you, and is listed as a h3 project. Are there plans in the next 90d to publish any code or examples or how Daemon is used by your PAI?
Because if it's a way to retrieve a Fact about you or another entity, it might solve a problem I haven't found an elegant solution for on my own, and from what I see and read, Daemon may be really useful and let me retrieve context and intents ad-hoc without necessitating constantly refactoring and carving into while trying to optimize my token utilization.
(My migration to 0.6.0 went great; it cost about $11. I would do it again, it was really entertaining and illuminating! But when I have back to back $5 days I suspect I am making a mistake somewhere.)