fix(core): add safe environment file rejection diagnostics

[sam2tom]

Problem and behavior

Environment file writes can return a rejection without a request-correlated reason in Core logs, making an upstream HTTP 400 insufficient to diagnose an upload failure.

Emit one api_request_rejected event for rejected requests that reach the environment file creation handler. Reuse the response writer and actual handler context so request IDs match X-Request-Id and trace/span IDs match traceparent. Log only the fixed operation/route, status, allowlisted code, and a finite reason classification; unclassified causes remain unknown. Do not log bodies, query strings, raw paths, resource IDs, arbitrary error messages, or credentials. HTTP responses, authorization, execution, and retry behavior are unchanged. Update the English and Chinese operations documentation together.

Validation

• Standard full API package tests passed.
• Focused affected API race tests and request/trace correlation tests passed.
• Documentation navigation, translation/hash, locale, and structure checks passed.
• A fresh independent review identified a trace-context omission; it was fixed and tested. A second fresh review found no actionable issues.
• Full API race testing hits an allocation-limit failure in TestUnknownMembersRejectWithLinearAllocation, reproduced against the unmodified baseline using a HEAD overlay. This unrelated failure is not reported as passing.

No deployment or paid runtime reproduction was performed. This diagnostic does not establish the historical upload failure's cause or claim that uploads have recovered.

[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 2026-10-07T14:50:41.118786Z	914e1d3	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.