Conversation
reverting this batch since the test is broken and needs to be looked at.
|
@codex review |
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 3f680be99f
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
@codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 6f1db1c42b
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| "*/api/generate*" | ||
| OR | ||
| "*/v1/chat/completions*" |
There was a problem hiding this comment.
The production rule's description says it monitors /api/chat, but the inspected SPL filters only /api/generate and the OpenAI-compatible /v1/chat/completions path. Native Ollama chat requests therefore never reach the extraction or threshold logic, so prompt-injection activity against the advertised /api/chat endpoint is missed.
Useful? React with 👍 / 👎.
| | rex field=_raw "\|\s+(?<status_code>\d+)\s+\|\s+(?<response_time>[\d\.]+[a-z]+)\s+\|\s+(?<src_ip>[\:\da-f\.]+)\s+\|\s+(?<http_method>\w+)\s+\"(?<uri_path>[^\"]+)\"" | ||
| | rex field=response_time "^(?:(?<minutes>\d+)m)?(?<seconds>[\d\.]+)s$" |
There was a problem hiding this comment.
For requests lasting at least a minute, Gin-style durations such as 1m2.3s cannot match this first rex because response_time permits only one numeric component followed by letters. Although the next rex explicitly handles minutes, it never receives these values, so the longest requests—well above the 30-second threshold—are silently excluded from the production detection.
Useful? React with 👍 / 👎.
| | where long_request_count > 55 | ||
| | eval avg_response_time=round(avg_response_time, 2) | ||
| | eval max_response_time=round(max_response_time, 2) | ||
| | eval severity=case(long_request_count > 50 OR max_response_time > 55, "critical", long_request_count > 20 OR max_response_time > 40, "high", 1=1, "medium" ) |
There was a problem hiding this comment.
Every emitted result has long_request_count > 55, so the immediately following long_request_count > 50 condition always assigns critical; the high and medium branches are unreachable regardless of response time. In production, even 56 requests that only slightly exceed 30 seconds are consequently mislabeled critical, which can drive disproportionate alert handling.
Useful? React with 👍 / 👎.
| modification_date: '2026-09-24' | ||
| author: Rod Soto | ||
| status: experimental | ||
| status: production |
There was a problem hiding this comment.
Promoting this rule exposes a production RCE finding for any single level=ERROR event containing broad terms such as model or server.go, because the inspected SPL only requires error_count > 0. Ordinary failures already listed under known_false_positives—for example an incompatible or corrupted model—therefore generate a medium-severity potential-RCE finding without any injection, traversal, or execution indicator.
Useful? React with 👍 / 👎.
|
Due to unexpected amount of changes required. Moving this to 6.9 |
This branch has not been deployed
1 participant
As discussed originally in #3710 - The content referenced in this PR was set to
experimentalintentionally to make it easier for us to test this with customers. Promoting them back.Note that the following rules were not in this scope, due to the fact that there is an issue currently with their test data / parsing. Hence they need to be looked at separately.
Keep in mind that this PR does not change logic in any way (where it does not make sense). Its simply a promotion PR.