agentleFS
Sign inStart free
ruifmruifm/aibox/AGENTS.md

aibox

AGENTS.md1 starsChanged 4 months agoMITView on GitHub

aibox is a lightweight Bash launcher that runs AI coding agents inside a Bubblewrap sandbox. It is the Nix/devShell-oriented successor to aidock: no container image, no session image commits, and no runtime configuration file. The public contract is intentionally small: - bind the current project read/write at the same absolute path; - keep host networking available; - expose Nix store and the Nix daemon socket; - mount only explicit agent state/config paths from $HOME; - mount fixed system agent config…

  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 31,843 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

  • In this repository2 other files it gives its agents.

The file

We split this AGENTS.md into the rules it gives an agent

7 rules. About 0 are common, 5 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    keep host networking available.

    line 12
  2. Distinctive:

    do not mount Git/GitHub config, SSH agent, GPG agent, the host D-Bus socket, Docker, or Kubernetes state.

    line 16
  3. Distinctive:

    Keep the runtime interface small.

    line 29
  4. Very specific:

    Keep Bash formatted with shfmt -i 4 -ci and ShellCheck-clean.

    line 31
  5. Very specific:

    Use Nix as the primary development path: nix develop -c just check.

    line 32
  6. Distinctive:

    Add tests for any mount-policy or CLI behavior change.

    line 33
  7. Distinctive:

    When behavior changes, update README.md in the same change.

    line 45

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

Agent Instructions

Project Overview

aibox is a lightweight Bash launcher that runs AI coding agents inside a Bubblewrap sandbox. It is the Nix/devShell-oriented successor to aidock: no container image, no session image commits, and no runtime configuration file.

The public contract is intentionally small:

  • bind the current project read/write at the same absolute path;
  • keep host networking available;
  • expose Nix store and the Nix daemon socket;
  • mount only explicit agent state/config paths from $HOME;
  • mount fixed system agent config directories read-only when present;
  • do not mount Git/GitHub config, SSH agent, GPG agent, the host D-Bus socket, Docker, or Kubernetes state;
  • expose selected host desktop services only through explicit flags and a filtered proxy;
  • print the final bwrap command to stderr before executing it.

--host-ro is an opt-in compatibility mode. It exposes host files read-only while retaining the workspace and selected state as writable. It keeps private runtime paths and does not expose the host systemd bus.

Development Rules

  • Edit aibox directly; there is no generated launcher.
  • Keep the runtime interface small. Do not add env-var configurability or config files. Use the existing selected-mount flags for explicit host paths.
  • Keep Bash formatted with shfmt -i 4 -ci and ShellCheck-clean.
  • Use Nix as the primary development path: nix develop -c just check.
  • Add tests for any mount-policy or CLI behavior change.

Testing

  • just check formats, lints, and runs unit tests.
  • just test-unit uses a fake bwrap and should not need kernel namespace support.
  • just test-integration uses real Bubblewrap and requires Linux user namespaces.

Documentation

When behavior changes, update README.md in the same change. The README should be explicit about prerequisites and the accident-safety threat model.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free