Skip to content

TR mirror drift in docs/usermanuals/en/data/audit.md: Audit is format-agnostic: it concatenates each row's text fields and runs the PII, secrets, near-duplicate, quality, language and leakage scans o… (DOC-TR12-001) #347

Open

Description

@cemililik

Wave 1 · theme D1-10: Documentation-only fixes · highest severity High · 1 finding(s) · effort docs fix

Summary

Neither. Only one text payload per row is scanned (pair join, else first of text/content/completion/prompt, else messages). DPO 'chosen'/'rejected' text is never scanned for PII/secrets.

Primary finding

DOC-TR12-001 — TR mirror drift in docs/usermanuals/en/data/audit.md: Audit is format-agnostic: it concatenates each row's text fields and runs the PII, secrets, near-duplicate, quality, language and leakage scans o…

severity High

Where. EN docs/usermanuals/en/data/audit.md:122 · TR docs/usermanuals/tr/data/audit.md:122

English text. Audit is format-agnostic: it concatenates each row's text fields and runs the PII, secrets, near-duplicate, quality, language and leakage scans over the result.

Turkish text. Audit format'tan bağımsızdır: her satırın metin alanlarını birleştirir ve sonuç üzerinde PII, sır, near-duplicate, kalite, dil ve sızıntı taramalarını çalıştırır.

Which is correct. Neither. Only one text payload per row is scanned (pair join, else first of text/content/completion/prompt, else messages). DPO 'chosen'/'rejected' text is never scanned for PII/secrets.

Proposed correction (Turkish text for the -tr page). Audit her satırdan tek bir metin yükü seçer: instruction+output, User+Assistant veya prompt+response çiftlerini birleştirir; yoksa sırayla text/content/completion/prompt sütunlarından ilkini, yoksa messages[].content değerlerini kullanır. Diğer alanlar (örn. DPO chosen/rejected) PII, sır, kalite, dil ve sızıntı taramalarının dışında kalır.

Acceptance criteria

  • Each statement is re-checked against the code at fix time.
  • Documentation listed above is corrected in English and in its Turkish mirror in the same PR.
  • Every finding in this issue is resolved or explicitly closed with a reason.

Found by the October 2026 ForgeLM code and documentation review against commit cab2947. Code links point at that commit; line numbers may have moved since. Finding IDs (CR-…, DOC-…, TECH-…) are stable references for this backlog.

Activity

  1. added
    documentationImprovements or additions to documentation
    i18nEnglish/Turkish documentation mirror
    severity: highWrong behaviour or a false claim users rely on
    source: review-2026-10Found by the October 2026 code and documentation review
    wave: 1Roadmap wave 1: high severity, small effort
    on Oct 9, 2026
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

    documentationImprovements or additions to documentationi18nEnglish/Turkish documentation mirrorseverity: highWrong behaviour or a false claim users rely onsource: review-2026-10Found by the October 2026 code and documentation reviewwave: 1Roadmap wave 1: high severity, small effort

    Type

    No type

    Projects

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions