Closed justinfenn closed 3 years ago
This sounds like a good addition. I can think of three different ways of implementing it:
value
optional for the encrypt
command, and when absent, pulls the data from stdin
.-
to tell Amber "grab the data from stdin"Problems with each of these that I see:
amber encrypt FOO
seems to work but does nothing-
for some reason, they can't (without extra workarounds)Overall I'm leaning towards (1). Any thoughts?
BTW, this is a very cool project! It hits a lot of sweet spots for in-repo secret storage.
Thank you!
I think 1 sounds good. If it prompts the user with something like "Secret value: " then it hopefully won't be too surprising. And it should still work if stdin is redirected to a file or something.
Could the encrypt subcommand take the secret value from stdin? This would help prevent raw secrets from being saved in shell history, for example.
BTW, this is a very cool project! It hits a lot of sweet spots for in-repo secret storage.