docs(codex): point GitLab readers to maintained guidance

[fwph-oai]

The older GitLab cookbook still uses the deprecated --full-auto flag and includes result handling that can convert invalid output to an empty findings array or continue after a nonzero Codex exit. Readers can also confuse its general code-quality/SAST-interpretation workflows with the dedicated Codex Security scanner.

This PR adds a source-visible note at the top of the existing article. It identifies the outdated flag and result-handling risk, links current non-interactive Codex guidance, and directs Codex Security scanning/remediation users to the maintained GitLab CI/CD guide introduced in openai/developers-website#2564.

The original examples remain unchanged. This is a scoped warning and routing correction, not a tested rewrite of the pipeline, a new archive designation, or a claim that the two workflows are interchangeable.

Johannes Bauer's review is requested on the distinction between general Codex interpretation of existing SAST findings and the dedicated Security workflow, and whether the maintained guide is the right destination for these readers.

Validation: repository Docs Editor review; verified current --full-auto deprecation and the dedicated GitLab guide; all added links resolve; exact comparison confirms all prior article text and code are preserved; git diff --check passed. No executable content changed and no CI pipeline or model call was run. Registry and author changes are not applicable.

Pre-merge docs-editor review rerun August 31 on 7e16f3186a386ce2d452a69bf41ad139c96d29e2: reviewed the notice in the full article context using the repository skill and style guide. No edits or new P0/P1 findings in the notice. Both added links return HTTP 200; the current non-interactive guide confirms the deprecated compatibility flag. Exact comparison confirms the entire original article and all four code fences remain unchanged, both image targets exist in the reviewed Git tree, and git diff --check passes. The legacy flag and failure/empty-output risks remain in the old examples and are explicitly documented above and in the note; this is not a validation or modernization of that pipeline. No notebook, registry, author, or executable changes.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-08-27T15:59:11.129019Z	a96e3c3	PR opened

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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[OAI-JB]

Looks good to me! Thanks for adding this to the archived Cookbook!