AI Decision System
Replayable, offline-verifiable records for AI-assisted go/no-go decisions: experiment promotions, typed model answers and human-approved actions.
FlatPark is a community hub for official Linux desktop apps — each one actually built, launched, and reviewed, then shipped as a signed, sandboxed Flatpak installed with one command.
Developer toolsShell · JavaScriptGitHub @flatparkOnlineOwner unconfirmed
Have you tried it? One tap is enough. No sign-in needed.
No updates detected yet. GitHub releases will appear here.
Owners of public repositories under a personal GitHub account can sign in to verify ownership and manage the introduction themselves. No administrator approval is needed.
Sign in, generate a verification file, and commit it to this repository’s default branch within 30 minutes. Files on other branches or forks do not count. Then check the file below to manage the project without administrator approval. Keep the file to retain your access.
If needed, you will go to sign-in. After signing in, return here and press your chosen verification button again.
Run /nomorevibe verify in your project folder to verify ownership and manage the introduction yourself.
Verify in one minuteAsk about other verification options
Yet another Flatpak hub.
Translation pending
Taken from the repository README
Loading the full README. The saved excerpt is shown below.
FlatPark is a Flatpak hub for apps that ship as a definitive download, built to
be driven by AI agents.Read the full README ↗FlatPark is a community Flatpak remote for Linux desktop apps that ship an
official download: a .deb, a tarball or an AppImage at a stable release URL.
Every app is the developer's own release, fetched from their servers at install
time and run in a sandbox. FlatPark doesn't build apps from source.
The goal is a remote that runs efficiently and is easy to take part in. Speed matters most: apps often land minutes after an upstream release, and a fix reaches users just as fast. Packaging, review, publishing and upgrades are built to be driven by AI agents. See About FlatPark and the principles.
flatpak --user remote-add --if-not-exists flatpark https://dl.flatpark.org/flatpark.flatpakrepo
flatpak --user install flatpark <app-id>
--user installs into your home (no admin); drop it from both commands for a
system-wide install. Browse the catalog at https://flatpark.org.
flatpak update.Use the apps. Install and use them. If something doesn't work, whether or not it has anything to do with Flatpak, open an issue and we'll handle it.
Request an app. If an app you want has no Flatpak, open an issue asking for it.
Lend your agent. If you have spare tokens, point your agent at
finding apps and let it search
for good apps that Flatpak users can't get yet. It files what it finds as
[app discover] issues.
Package an app. New apps are very welcome. Pick one from the issues or bring your own. Open this repo in your agent (e.g. Claude Code) and ask:
I want to publish <app or upstream URL> to FlatPark. Start from
docs/README.mdand follow the packaging playbook.
The packaging playbook takes it from manifest to tested PR.
Everything else. PRs for the website, the documentation and translations are just as welcome, and so is any other improvement.
See CONTRIBUTING.md and the documentation.
If your app is on FlatPark, you can link to it from your README or website with this badge:
Markdown (replace <app-id>):
[](https://flatpark.org/apps/<app-id>)
HTML:
<a href="https://flatpark.org/apps/<app-id>">
<img alt="Get it on FlatPark" height="56" src="https://flatpark.org/badges/get-it-on-flatpark.svg">
</a>
A PNG (2×, 308×112) is at https://flatpark.org/badges/get-it-on-flatpark.png for places that don't render SVG.
FlatPark's own code is MIT. It does not license the packaged applications themselves — those remain their vendors' property and are fetched from official sources at install time.