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-4949GitHub issuePublicReceived

omarchy-agent-usage-codex passes a removed --ask-for-approval value, so Codex limits never load

Reported by CougarM · filed Sep 7, 2026Closed upstream on 2026-09-20 by bjarneo via an automated message, as a duplicate of #8460 (R-1406). GitHub close reason: completed.

Received and recorded verbatim. Nothing has been checked yet.

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?
Not yet mapped to any requirement or component.
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?
Our staff decide whether the report is valid. Where this project allows it, agents decide whether it duplicates another report or a known issue. No decision has been recorded yet.

Original report

omarchy-agent-usage-codex passes a removed --ask-for-approval value, so Codex limits never loadrevision 1 · captured 2026-10-04 12:10 UTC via poll

Title: omarchy-agent-usage-codex passes a removed --ask-for-approval value, so Codex limits never load

What happens

The Agents bar widget always shows "Codex limits unavailable" for Codex, with an empty limits[], even with a fully authenticated codex CLI. Claude limits work fine on the same machine.

The written record ends up as:

{
  "id": "codex",
  "ready": true,
  "tierLabel": "",
  "usageStatusText": "Codex limits unavailable",
  "authHelpText": "initialize",
  "limits": []
}

Cause

bin/omarchy-agent-usage-codex:531 starts the app-server with -a untrusted:

proc = subprocess.Popen(
  [codex, "-s", "read-only", "-a", "untrusted", "app-server"],

untrusted is no longer a valid --ask-for-approval value in current Codex CLI (0.153.4):

$ codex -s read-only -a untrusted app-server
error: invalid value 'untrusted' for '--ask-for-approval <APPROVAL_POLICY>'
  [possible values: on-request, never]

The process therefore exits immediately, rpc_request never sees a reply, and its
raise TimeoutError(method) surfaces as authHelpText: "initialize" — which reads like an
auth or handshake problem rather than a bad CLI argument.

Fix

-a never works. Same sequence against the same CLI, stdin held open between calls:

{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"clientInfo":{"name":"probe","version":"1"}}}
{"jsonrpc":"2.0","id":2,"method":"account/read","params":{}}
{"jsonrpc":"2.0","id":3,"method":"account/rateLimits/read","params":{}}

returns exactly what the collector wants:

{"id":2,"result":{"account":{"type":"chatgpt","planType":"<plan>"}}}
{"id":3,"result":{"rateLimits":{
  "primary":  {"usedPercent":<n>,"windowDurationMins":300,  "resetsAt":<epoch>},
  "secondary":{"usedPercent":<n>,"windowDurationMins":10080,"resetsAt":<epoch>}}}}

(values redacted; windowDurationMins 300 / 10080 are the 5-hour and weekly windows
the collector maps to its two limit rows.)

So the one-line change is "untrusted" → "never" on line 531.

Secondary suggestion

stderr=subprocess.DEVNULL on that Popen is what made this hard to diagnose — the CLI
printed a precise error and it was discarded, leaving only a timeout on initialize.
Capturing stderr and putting it in authHelpText when the handshake fails would have
made this self-evident. Worth considering independently of the fix above, since it will
happen again the next time the Codex CLI changes an argument.

Steps to reproduce

  1. Authenticate the Codex CLI (codex login) on a ChatGPT plan.
  2. Run omarchy-agent-usage-codex --force.
  3. Inspect ~/.local/state/omarchy/agents/usage/codex.json — limits is [] and

authHelpText is "initialize".
4. Run codex -s read-only -a untrusted app-server directly to see the argument error.

System

  • Omarchy 4.0.1-1
  • Codex CLI 0.153.4
  • Arch Linux, Hyprland 0.56.2
View raw text
**Title:** `omarchy-agent-usage-codex` passes a removed `--ask-for-approval` value, so Codex limits never load

## What happens

The Agents bar widget always shows **"Codex limits unavailable"** for Codex, with an empty `limits[]`, even with a fully authenticated `codex` CLI. Claude limits work fine on the same machine.

The written record ends up as:

```json
{
  "id": "codex",
  "ready": true,
  "tierLabel": "",
  "usageStatusText": "Codex limits unavailable",
  "authHelpText": "initialize",
  "limits": []
}
```

## Cause

`bin/omarchy-agent-usage-codex:531` starts the app-server with `-a untrusted`:

```python
proc = subprocess.Popen(
  [codex, "-s", "read-only", "-a", "untrusted", "app-server"],
```

`untrusted` is no longer a valid `--ask-for-approval` value in current Codex CLI (0.153.4):

```
$ codex -s read-only -a untrusted app-server
error: invalid value 'untrusted' for '--ask-for-approval <APPROVAL_POLICY>'
  [possible values: on-request, never]
```

The process therefore exits immediately, `rpc_request` never sees a reply, and its
`raise TimeoutError(method)` surfaces as `authHelpText: "initialize"` — which reads like an
auth or handshake problem rather than a bad CLI argument.

## Fix

`-a never` works. Same sequence against the same CLI, stdin held open between calls:

```
{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"clientInfo":{"name":"probe","version":"1"}}}
{"jsonrpc":"2.0","id":2,"method":"account/read","params":{}}
{"jsonrpc":"2.0","id":3,"method":"account/rateLimits/read","params":{}}
```

returns exactly what the collector wants:

```json
{"id":2,"result":{"account":{"type":"chatgpt","planType":"<plan>"}}}
{"id":3,"result":{"rateLimits":{
  "primary":  {"usedPercent":<n>,"windowDurationMins":300,  "resetsAt":<epoch>},
  "secondary":{"usedPercent":<n>,"windowDurationMins":10080,"resetsAt":<epoch>}}}}
```

(values redacted; `windowDurationMins` 300 / 10080 are the 5-hour and weekly windows
the collector maps to its two limit rows.)

So the one-line change is `"untrusted"` → `"never"` on line 531.

## Secondary suggestion

`stderr=subprocess.DEVNULL` on that `Popen` is what made this hard to diagnose — the CLI
printed a precise error and it was discarded, leaving only a timeout on `initialize`.
Capturing stderr and putting it in `authHelpText` when the handshake fails would have
made this self-evident. Worth considering independently of the fix above, since it will
happen again the next time the Codex CLI changes an argument.

## Steps to reproduce

1. Authenticate the Codex CLI (`codex login`) on a ChatGPT plan.
2. Run `omarchy-agent-usage-codex --force`.
3. Inspect `~/.local/state/omarchy/agents/usage/codex.json` — `limits` is `[]` and
   `authHelpText` is `"initialize"`.
4. Run `codex -s read-only -a untrusted app-server` directly to see the argument error.

## System

- Omarchy 4.0.1-1
- Codex CLI 0.153.4
- Arch Linux, Hyprland 0.56.2

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

Not linked to any known issue, other report or change yet.

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.

No decision has been recorded yet.