Skip to content

Scope clarification for reported cybersecurity safeguard interruptions #54671

Open

Description

@leogranda

What version of Codex CLI is running?

codex-cli 0.163.0 — current version; version during the October 14 incidents unknown.

What subscription do you have?

Pro

Which model were you using?

gpt 6.1 Sol xhigh

What platform is your computer?

Linux 7.0.0-34-generic x86_64 unknown — current host output; historical kernel version not verified.

What terminal emulator and version are you using (if applicable)?

Remote Linux host accessed over SSH. Terminal emulator version and multiplexer use during the incidents are not recorded.

Codex doctor report

Not collected for this report.

What issue are you seeing?

The assistant’s conversation history reported two interruptions with the message:

“This content was flagged for possible cybersecurity risk”

These reports appeared in messages dated October 14, 2026, at 15:24:36 and 16:06:36 UTC, during preparation/review of local resource-cleanup tests for
our own audit service.

Those are message timestamps, not verified tool-event timestamps. We do not have the original tool notices, exact blocked requests, or request IDs. We
cannot establish whether this was a false positive, which component generated the notice, or whether the restriction remains applicable.

We are requesting investigation and guidance on the appropriate support channel, not a bypass or removal of safeguards.

What steps can reproduce the bug?

Uploaded thread: 01a0c5a5-9a00-79d0-97f1-ae364103b821

What is the expected behavior?

An actionable diagnostic identifying the affected component and, where available, a request reference and the appropriate review/support process.

Please clarify how we can determine whether the proposed local test scope is permitted, or which private support channel should handle that question.

We do not expect operator authorization or a feedback submission to override platform safeguards.

Additional information

The pending tests concern resource cleanup, file-descriptor ownership, and commit/ACK/replay behavior in a service we own. They use synthetic data,
local AF_UNIX sockets, and processes we create, without sudo or production changes.

The current proposed profile does not use strace or ptrace. It has not been executed.

A historical kernel ptrace denial and later AppServerBackendRequestError failures are separate incidents. We have no evidence that they share a cause
with the reported cybersecurity notices.

No credentials, private logs, or production databases are included in this public report. Please indicate a private channel if further sensitive
diagnostics are required.

Activity

  1. added
    CLIIssues related to the Codex CLI
    safety-checkIssues related to safety and abuse checks
    on Oct 16, 2026
  2. leogranda commented on Oct 16, 2026

    @leogranda
    Author
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    CLIIssues related to the Codex CLIbugSomething isn't workingsafety-checkIssues related to safety and abuse checks

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions