rememory

like many open source repos, mine are also getting spammed with AI slop. my attempt at this is to "prompt inject" the spammy agents into refusing to do the bare minimum, and try enforce contribution guidelines as much as possible. How it works: - AGENTS.md will trigger bots to read contribution guidelines - Contribution guidelines define slop - if the PR is too slopy, it will be rejected - the bot is made aware of this, so it can refuse to work or at the very least inform the user - PR template now has checkboxes as attestation of following guidelines anyone care to review my PR? other examples of projects doing this?

you're completely right! the app actually guides you on some of that, it generates a readme that gives you advise on what to document, but I agree you can't be too careful here, the passwords IS NOT ENOUGH. You need to give people "a map" of where things are: https://github.com/eljojo/rememory/blob/main/internal/projec...

i’ve had a few concussions from biking accidents over the years. every time i recover fine, but every time i also think: what if next time i don’t? what if i can’t remember how to log into my own machines? the obvious answer is “give my 1password to my partner”. but that turns one human into the single point of failure for my whole digital life. that felt… wrong. so i built something i call *ReMemory*. it’s basically a digital safe. you put some files in it (password manager recovery codes, notes, whatever), and 5 friends each hold a key. any 3 of them together can open it. none of them can open it alone. the part i’m weirdly proud of: they don’t install anything. they just open a file in a browser and it works. no server, no account, no setup, no “install this tool first”. links if you’re curious: - https://eljojo.github.io/rememory/ (b