False positive: authorized local code review blocked as cybersecurity request
Nobody has claimed this yet.
- Dominant language
- Rust
- Stars
- 125k
- Forks
- 19.4k
- PR merge metrics
- PR metrics pending
Description
What version of Codex CLI is running?
codex-cli 0.153.4
What subscription do you have?
ChatGPT Plus
Which model were you using?
gpt-5.6-sol, reasoning effort: max
What platform is your computer?
Windows 11
What terminal emulator and version are you using (if applicable)?
Wrap
Codex doctor report
{
"schemaVersion": 1,
"overallStatus": "warning",
"codexVersion": "0.153.4",
"environment": {
"platform": "windows-x86_64",
"os": "Windows 11 Professional 64-bit, build 26200",
"terminal": "Warp",
"installMethod": "standalone",
"authMode": "chatgpt",
"storedApiKey": false
},
"checks": {
"installation": {
"status": "ok",
"summary": "installation looks consistent"
},
"updates": {
"status": "ok",
"summary": "current version is not older",
"latestVersion": "0.153.4"
},
"networkHttp": {
"status": "ok",
"summary": "active provider endpoints are reachable"
},
"networkWebsocket": {
"status": "ok",
"summary": "Responses WebSocket handshake succeeded"
},
"config": {
"status": "ok",
"summary": "configuration loaded successfully"
},
"localSecurityEnforcement": {
"status": "ok",
"summary": "no locally visible recent Codex security enforcement was found"
},
"endpointProtection": {
"status": "warning",
"summary": "endpoint protection detected; Codex exclusions are unverified"
},
"threadInventory": {
"status": "warning",
"summary": "2 rollout files are missing from the state DB and 1 duplicate thread identifier was detected"
},
"terminalWidth": {
"status": "warning",
"summary": "terminal width is 65 columns; 80 or more is recommended"
}
},
"uploadedFeedbackThreadId": "01a07921-78ff-73e2-81b0-ad48311e8845"
}
What issue are you seeing?
Description
Codex incorrectly blocked an authorized defensive review of my own local software repository as a restricted cybersecurity request.
The requested work was limited to reviewing a newly implemented HTTP endpoint against the repository's existing design contract and automated tests. All testing was explicitly restricted to localhost, synthetic data, temporary SQLite storage, and systems under my control.
The prompt did not request access to external systems, credentials, persistence, exploitation, malware, or unauthorized activity.
Steps to reproduce
- Open Codex CLI in a local repository owned by the user.
- Select
gpt-5.6-solwith reasoning effortmax. - Request a read-only review of a local HTTP endpoint and its automated tests.
- Restrict all checks to localhost, synthetic data, and temporary storage.
- Codex displays: “This content can't be shown” and redirects to Trusted Access.
Expected behavior
Codex should recognize the request as ordinary authorized defensive code review, or ask the user to narrow the scope if any individual check is considered too sensitive.
Actual behavior
The entire request is blocked before the repository review begins.
Diagnostics
Session and logs were uploaded from Codex CLI.
Thread ID: 01a07921-78ff-73e2-81b0-ad48311e8845
No credentials, production systems, external targets, or real personal data were involved.
What steps can reproduce the bug?
Uploaded thread: 01a07921-78ff-73e2-81b0-ad48311e8845
What is the expected behavior?
No response
Additional information
No response
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the uploaded feedback thread 01a07921-78ff-73e2-81b0-ad48311e8845 and its session logs to trace why the localhost-only repository review was classified as restricted. Reproduce the request described in the issue and compare the resulting block with the expected authorized-review behavior. Done means the review is allowed or the user receives a narrower-scope request instead of a blanket block.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- rust
- Domain
- cli, security
- Issue type
- Bug
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 42/100