🚀 Feature: ephemeral private key (do not store private key on disk) #304

New Issue

OVERLORD · 2026-02-04T18:42:16+03:00

OVERLORD commented

2026-02-04 18:42:16 +03:00

Originally created by @ItalyPaleAle on GitHub (May 28, 2025).

Feature description

Instead of saving the private key to disk, Pocket ID could optionally generate a key when it starts up and keep it in-memory only. This would be optional.

Pros:

Private keys are never stored un-encrypted on disk (see also #580)
The key is rotated frequently, every time Pocket ID starts up

Cons:

The key changes every time Pocket ID is started, which means that tokens issued by Pocket ID would be invalidated on every restart. This may or may not be acceptable depending on what the downstream clients expect. This is why this feature is optional.

Pitch

This is another relatively simple feature to implement that helps protecting the "keys to the kingdom". It may not be suitable for all scenarios, but many users who run Pocket ID in their homelab would likely have minimal to no impact when using this.

Originally created by @ItalyPaleAle on GitHub (May 28, 2025). ### Feature description Instead of saving the private key to disk, Pocket ID could optionally generate a key when it starts up and keep it in-memory only. This would be **optional**. Pros: - Private keys are never stored un-encrypted on disk (see also #580) - The key is rotated frequently, every time Pocket ID starts up Cons: - The key changes every time Pocket ID is started, which means that tokens issued by Pocket ID would be invalidated on every restart. This _may or may not_ be acceptable depending on what the downstream clients expect. This is why this feature is optional. ### Pitch This is another relatively simple feature to implement that helps protecting the "keys to the kingdom". It may not be suitable for _all_ scenarios, but many users who run Pocket ID in their homelab would likely have minimal to no impact when using this.

OVERLORD added the feature label 2026-02-04 18:42:16 +03:00

OVERLORD closed this issue

2026-02-04 18:42:22 +03:00

OVERLORD commented

2026-02-04 18:42:30 +03:00

@kmendell commented on GitHub (May 28, 2025):

I think the encryption idea makes sense, but this would introduce a point of failure and may just cause confusion for people if they are not the best at troubelshooting as well.

@kmendell commented on GitHub (May 28, 2025): I think the encryption idea makes sense, but this would introduce a point of failure and may just cause confusion for people if they are not the best at troubelshooting as well.

OVERLORD commented

2026-02-04 18:42:39 +03:00

@ItalyPaleAle commented on GitHub (May 28, 2025):

Why would it be a point of failure?

Agree it can be confusing. I am thinking this should be optional for this reason.

@ItalyPaleAle commented on GitHub (May 28, 2025): Why would it be a point of failure? Agree it can be confusing. I am thinking this should be optional for this reason.

OVERLORD commented

2026-02-04 18:42:48 +03:00

@kmendell commented on GitHub (May 29, 2025):

Im closing this as we can just track it in https://github.com/pocket-id/pocket-id/issues/580, since its seems like it can be part of that.

@kmendell commented on GitHub (May 29, 2025): Im closing this as we can just track it in https://github.com/pocket-id/pocket-id/issues/580, since its seems like it can be part of that.

Sign in to join this conversation.

Branches Tags

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: starred/pocket-id#304