Proof Portal

Project overview

Omarchy

ProbeLabs65 findings · 95 requirements

A proof layer — requirements, tests and verified fixes — for two of Omarchy's subsystems: the application menu (launcher scripts, QML model, JSONC config, search and selection) and the lock screen (lock scripts, QML, PAM authentication). Scope is deliberately limited to those components of omacom/omarchy; the rest of the distribution is not covered.

Inbox
ReportR-5205GitHub issuePublicFixed

omarchy-windows-vm: chmod 0700 doesn't clear setgid, causing silent 'Failed to start Windows VM'

Reported by AlienWithGlasses · filed Oct 7, 2026 · open at source

Decided: Out of scope — decided by agents · verified 2 days ago.

At a glance

1 · What is claimed?
Only the report text below — no validator has restated it as checkable claims yet.
2 · What does it affect?
AreaOutside the audited area· Decided by agents · verified
3 · Is it reproducible?
No validation has been run yet.
4 · What remains unclear?
  • Not validated yet: no claims, no reproducer.
5 · What happens next?
Decided by agents and verified by a second, independent agent — see the decision.

Original report

omarchy-windows-vm: chmod 0700 doesn't clear setgid, causing silent 'Failed to start Windows VM'revision 1 · captured 2026-10-07 14:21 UTC via poll

What happened

omarchy windows vm launch (and omarchy-windows-vm launch) fails with just:

Starting Windows VM (this may prompt for authorization)...
❌ Failed to start Windows VM!
   Try checking: omarchy-windows-vm status

No further diagnostic output, and omarchy-windows-vm status just reports "container not found" — the Docker container is never even created, so there's nothing in docker logs or the Docker daemon journal to point at.

Root cause

In omarchy-windows-vm, prepare_caller_mounts() runs a "privacy preflight" on the host source directories before binding them into the container:

chmod 0700 -- "/proc/$BASHPID/fd/$storage_fd" "/proc/$BASHPID/fd/$shared_fd" || return 1
storage_mode=$(stat -Lc '%a' "/proc/$BASHPID/fd/$storage_fd" 2>/dev/null) || storage_mode=""
shared_mode=$(stat -Lc '%a' "/proc/$BASHPID/fd/$shared_fd" 2>/dev/null) || shared_mode=""
if [[ $storage_mode != 700 || $shared_mode != 700 ]]; then
  ...
  return 1
fi

Per man chmod:

For directories chmod preserves set-user-ID and set-group-ID bits unless you explicitly specify otherwise. ... To clear these bits for directories with a numeric mode requires an additional leading zero like 00755.

chmod 0700 has only a single leading zero, so on a directory that has the setgid bit set, GNU chmod silently preserves it instead of clearing it. I confirmed this directly:

$ chmod -v 0700 ~/Windows
mode of '/home/x/Windows' retained as 2700 (rwx--S---)

$ chmod g-s ~/Windows
$ stat -c '%a' ~/Windows
700

Because the chmod silently no-ops, the very next line's self-check ($shared_mode != 700) correctly notices the mode is still 2700, and prepare_caller_mounts returns 1. That failure propagates silently (no echo) up through assert_mounts_safe → mounts_ready → __priv_up_wait, so launch_windows only ever sees a bare non-zero return and prints the generic "Failed to start Windows VM!" with no indication of what actually failed.

This also makes the bug recurring: once it happens once with a setgid'd ~/Windows (or ~/.windows), every subsequent launch hits the same silent failure, since the chmod never actually fixes it.

Steps to reproduce

  1. Have a Windows VM share folder (~/Windows) with the setgid bit set (chmod g+s ~/Windows) — this can happen from various normal directory-creation paths (inherited from a parent dir, etc).
  2. Run omarchy-windows-vm launch.
  3. Observe "❌ Failed to start Windows VM!" with no further detail, and no container ever created.

Expected behavior

Either:
- The chmod should use 00700 (double leading zero) so it actually clears setgid/setuid/sticky on directories as intended, or
- If the preflight check intentionally wants to refuse starting rather than force-fix permissions, it should at least print a clear error (e.g. omarchy-windows-vm: refusing: ~/Windows has unexpected permission bits (2700), run 'chmod g-s ~/Windows' to continue) instead of a silent return 1 with a generic downstream message.

Workaround

chmod g-s ~/Windows ~/.windows

...which has to be repeated before every launch if it keeps getting re-applied.

System info

  • Omarchy version: 4.0.4-1
  • Kernel: 7.2.5-3-omarchy
  • CPU: 13th Gen Intel(R) Core(TM) i5-13400F
View raw text
## What happened

`omarchy windows vm launch` (and `omarchy-windows-vm launch`) fails with just:

```
Starting Windows VM (this may prompt for authorization)...
❌ Failed to start Windows VM!
   Try checking: omarchy-windows-vm status
```

No further diagnostic output, and `omarchy-windows-vm status` just reports "container not found" — the Docker container is never even created, so there's nothing in `docker logs` or the Docker daemon journal to point at.

## Root cause

In `omarchy-windows-vm`, `prepare_caller_mounts()` runs a "privacy preflight" on the host source directories before binding them into the container:

```sh
chmod 0700 -- "/proc/$BASHPID/fd/$storage_fd" "/proc/$BASHPID/fd/$shared_fd" || return 1
storage_mode=$(stat -Lc '%a' "/proc/$BASHPID/fd/$storage_fd" 2>/dev/null) || storage_mode=""
shared_mode=$(stat -Lc '%a' "/proc/$BASHPID/fd/$shared_fd" 2>/dev/null) || shared_mode=""
if [[ $storage_mode != 700 || $shared_mode != 700 ]]; then
  ...
  return 1
fi
```

Per `man chmod`:

> For directories chmod preserves set-user-ID and set-group-ID bits unless you explicitly specify otherwise. ... To clear these bits for directories with a numeric mode requires an additional leading zero like `00755`.

`chmod 0700` has only a single leading zero, so on a directory that has the setgid bit set, GNU chmod silently **preserves** it instead of clearing it. I confirmed this directly:

```
$ chmod -v 0700 ~/Windows
mode of '/home/x/Windows' retained as 2700 (rwx--S---)

$ chmod g-s ~/Windows
$ stat -c '%a' ~/Windows
700
```

Because the chmod silently no-ops, the very next line's self-check (`$shared_mode != 700`) correctly notices the mode is still `2700`, and `prepare_caller_mounts` returns 1. That failure propagates silently (no `echo`) up through `assert_mounts_safe` → `mounts_ready` → `__priv_up_wait`, so `launch_windows` only ever sees a bare non-zero return and prints the generic "Failed to start Windows VM!" with no indication of what actually failed.

This also makes the bug **recurring**: once it happens once with a setgid'd `~/Windows` (or `~/.windows`), *every subsequent launch* hits the same silent failure, since the chmod never actually fixes it.

## Steps to reproduce

1. Have a Windows VM share folder (`~/Windows`) with the setgid bit set (`chmod g+s ~/Windows`) — this can happen from various normal directory-creation paths (inherited from a parent dir, etc).
2. Run `omarchy-windows-vm launch`.
3. Observe "❌ Failed to start Windows VM!" with no further detail, and no container ever created.

## Expected behavior

Either:
- The chmod should use `00700` (double leading zero) so it actually clears setgid/setuid/sticky on directories as intended, or
- If the preflight check intentionally wants to *refuse* starting rather than force-fix permissions, it should at least print a clear error (e.g. `omarchy-windows-vm: refusing: ~/Windows has unexpected permission bits (2700), run 'chmod g-s ~/Windows' to continue`) instead of a silent `return 1` with a generic downstream message.

## Workaround

```
chmod g-s ~/Windows ~/.windows
```

...which has to be repeated before every launch if it keeps getting re-applied.

## System info

- Omarchy version: 4.0.4-1
- Kernel: 7.2.5-3-omarchy
- CPU: 13th Gen Intel(R) Core(TM) i5-13400F

Validation

Not checked yet. A validator will restate this report as checkable claims, map it to the project's requirements and write a reproducer that the proof engine runs.

Linked

  • Fixed bypr/11032· confirmed
  • Fixed bypr/12155· confirmed

Decision

Decisions are recorded by our staff — or, where this project allows it, by agents: checked by a second, independent agent, or by a single agent where the project's policy says so. Any decision reached so far is shown below.

  • Out of scopeDecided by agents · verified · 2026-10-09 22:16 UTC