• peanut-walrus 4 hours ago

    The main problems with these kinds of in-repo vault solutions:

    - Sharing encryption key for all team members. You need to be able to remove/add people with access. Only way is to rotate the key and only let the current set of people know about the new one.

    - Version control is pointless, you just see that the vault changed, no hint as to what was actually updated in the vault.

    - Unless you are really careful, just one time forgetting to encrypt the vault when committing changes means you need to rotate all your secrets.

    • nothrabannosir 3 hours ago

      Agreed with 1 and 3, just a tip re 2 though: sops encodes json and yaml semantically, key names of objects are preserved. Iow you can see which key changed.

      Whether that is a feature or a metadata leak is up to the beholder :)

    • mbreese 5 hours ago

      Secrets management is hard. And proper secret sharing setups meant for larger groups are quite unwieldy to work with with smaller groups. Well, they are hard to work with for all sizes of groups, but it seems particularly overkill for small groups. So I see why you'd want to do this. I also kinda like the idea of just encrypting/decrypting .env files. It's a pretty clean design.

      But storing secrets in the same git repository just seems off to me. I don't like the idea of keeping the secrets (even in encrypted form) with the code I'm deploying.

      There should be a better balance somewhere, but I'm not sure this is quite it for me. Shared keepass files (not in git) or 1Password vaults are harder to work with, but I think lean more towards the secure side at the expense of a bit of usability. (Depending on the team, OSs, etc...)

    • e12e 4 hours ago

      This looks nice, but I think I lean towards fnox (by the author of mise) - because of the flexibility and support for external storage:

      https://github.com/jdx/fnox

      • mrinterweb 2 hours ago

        This reminds me of how rails manages encrypted credentials. https://guides.rubyonrails.org/security.html#custom-credenti.... The big differences, this is portable and sets values as session env vars instead of application vars.

        • steffoz 6 hours ago

          Very similar to a tool I built about a year ago! We've been using it with our 6-person team, and it's been working great. It uses a shared keyring.json to manage public keys, so we don’t have to duplicate the same keys across every repo.

          https://github.com/stefanoverna/kavo

          It’s built on top of age for encryption (https://github.com/FiloSottile/age).

          • andreineculau 5 hours ago

            I understand the simplicity angle. https://github.com/elasticdog/transcrypt has been around for a long time and strikes that balance very well in my opinion. And it's just a bash script that can also be committed so the git repo is atomic.

            • madeforhnyo 2 hours ago

              Being a node dev - by necessity, I've settled on dotenvx [0] for committing encrypted .env files.

              [0] https://dotenvx.com/

              • n31l 35 minutes ago

                Agreed, and it's nice and easy for anyone already using `.env` files, although the private key used to decrypt the dotenvx key-values is itself a secret.

                • hersko 20 minutes ago

                  Yeah i don't understand this. You still need to secure your .env.keys file same as you would be doing with a standard .env. Is the benefit just that you can track it with git?

              • eddyg 4 hours ago

                I’ve been using git-crypt⁽¹⁾ which is transparent (you put the patterns you want to encrypt in .gitattributes) and lets you use GPG keys or symmetric keys. And it's been around for quite a while.

                ⁽¹⁾https://github.com/AGWA/git-crypt

                • jillesvangurp 9 hours ago

                  Sounds useful. We do similar things with encrypted properties files. Also, things like Ansible come with ansible vault. If you use Github, you can use Github secrets of course. And AWS/GCP/etc. tend to have secret stores.

                  The challenge with this solution is of course managing who has access and dealing with people leaving your team and no longer being trusted. Even if you still like them personally, just because they are outside your team would require you to change any credentials they might have.

                  In our case, our team is small and I simply ignore this problem. So, we have a keepass file with shared secrets and repositories with encrypted properties files and a master password in this keepass file. Mostly, it's just me handling the password. It also gets configured as a Github secret on repositories for CI and deployment jobs. It works. But I'm aware of the limitations.

                  This is an area where there are lots of tools but not a whole lot of standardized ones or good practices for using them. It's one of those things that acts as a magnet for enterprise complexity. Tools like this tend to become very unwieldy because of this. Which is why people keep reinventing them.

                  • crote 8 hours ago

                    > The challenge with this solution is of course managing who has access and dealing with people leaving your team and no longer being trusted. Even if you still like them personally, just because they are outside your team would require you to change any credentials they might have.

                    At least it's a clearly exposed problem: everyone who has ever cloned the repo has a copy of your secrets.

                    With software like 1Password it is way too easy to blindly rely on built-in permission management. People implicitly assume that removing a person's 1Password access means they can no longer rely the underlying resource - but in practice they could've copied the secret onto a sticky note at any time, and it's not safe until you've rotated the secret!

                    With shared user accounts there's at least usually the possibility of using 2FA - but that's not exactly going to work with things like deployment tokens intended for automated use...

                    Of course in an ideal world we wouldn't have those kinds of secrets and we'd all be using short-lived tightly-scoped service accounts - but we don't live in an ideal world.

                    • pverheggen an hour ago

                      Regarding the sticky note problem, this can be mitigated with separate vault credentials for production. That way you can limit prod secrets to a much smaller group, and if you wanted to rotate when someone leaves, you'd have to do it much less often.

                    • shoemann 9 hours ago

                      Absolutely agree. That is exactly why I made this tool - my projects usually don't have ansible, github, aws and other external dependencies, or have different sets of such dependencies, and teams are too small to use something enterprise level.

                    • submain 7 hours ago

                      This is great! Coincidentally, I just started replacing my collection of bespoke security bash scripts with an app like yours. WIP here: https://github.com/leolimasa/age-vault

                      We all keep reinventing the same thing :)

                      • sureglymop 4 hours ago

                        I understand the thought but my honest advice is: do not commit secrets to git, even if they are encrypted.

                        Secrets are not configuration, they are state (and I would say, an even stricter form of state that should ideally only exist at runtime in memory).

                        • Barathkanna 8 hours ago

                          This actually looks handy for the “small team with a couple of env files” use case. Most secret-management tools are great once you’re at scale, but trying to explain sops or git-crypt to a team that just wants to stop pasting secrets into Slack is… not fun. A simple password-protected vault committed to git is a reasonable middle ground.

                          I like the OS keyring integration too,removes a lot of friction. Curious how it behaves in multi-machine workflows and whether you plan to add any guardrails around accidental plaintext commits, since that’s usually where lightweight tools get tripped up.

                          • 8cvor6j844qw_d6 8 hours ago

                            > stop pasting secrets into Slack

                            You got me interested. I've seen sharing of API keys via Discords in hackathons.

                        • chrisweekly 5 hours ago

                          Committing the vault to git gives me the heebie-jeebies. (Not that I have a better solution with anything like this convenience.)

                          • sshine 5 hours ago

                            The way I think about it is:

                              - Maintaining stateful secrets at rest gives me the heebie-jeebies.
                              - The tools shouldn't let me shoot myself in the foot.
                              - The tools should ideally not have such a high learning curve that I won't actually use them.
                            
                            You can put your secrets in a separate repository and not think of them as the same kind of repository you'd publish.

                            Like... I wouldn't put a git-crypt'ed / sops-nix'ed repository online, simply because I don't like the idea that now anyone needs is brute-force; I know quantum computers aren't there yet wrt. brute-forcing stuff made by random people like me, but even hypothetically having this attack vector open, I don't like it.

                            So there's only two good solutions:

                              - You put secrets in a (hashicorp-style) vault that only decrypts temporarily in memory.
                              - You put secrets in an encrypted database with only safe tool integration.
                            
                            The things I don't like about git-based secrets management:

                              1. You might mix your secrets into projects and then later someone else might release that (against your current interest)
                              2. The solutions I've seen (sops-nix, agenix, secrix, etc.) are hard to set up and even harder to onboard people on
                            
                            When something's hard to set up, you might make a mistake or skip some concept.

                            Well-done secrets management that isn't based on a service like AWS Secrets og GitHub Secrets should be much, much easier.

                            I like the idea of how easy this is. Now, if it would just be best practice in every possible way at the same time.

                            The (admittedly well-known) problem with lockenv is that you can't revoke access once a password is known.

                            It's a big ask.

                          • akabalanza 8 hours ago

                            That looks amazing, thanks for sharing!

                            I have a git-based sync tool for my dotenv files. Maybe I can store my ssh keys, too

                            • rcarmo 9 hours ago

                              I use a Makefile target with GPG :)