fix(report): harden generic diagnostics title boundary

[seonghobae]

Closes #999.

Direct child of #1002 at exact parent 52c7401bdec44a7a1e6194f0336617bfc5fe6eed; merge order is #1002 then #1000.

TDD lineage

• RED 18137218ad3eb825a5b792cb3dd1cf8f568b1b62: public render_diagnostics_report() must reject a caller-defined str subclass before overridden __bool__ or replace() can execute, raise package-owned ValueError("title must be a string or None"), and create no output artifact. It also preserves the empty exact built-in string → default-title contract.
• GREEN 54b374c9ebbd8412e2f7cf51bc6cc7684b1ff8b1: admit only None or an exact built-in str before truth-value testing and HTML escaping.
• Evidence e58df9ae468f676e332616581f26a464cddd7e05: governed changelog fragment.
• Strengthening 4d34e34e667109598eed7cb44555a6a072690681: prove an admitted exact built-in title remains escaped in both <title> and <h1> surfaces.
• Stack sync 528d980c1613fd6326f4c5c59effc85e80fb98b2: record current fix(report): replay main focus accessibility on current review workflow #1002 parent ancestry after both branches restored protected-main's exact .Jules/palette.md blob.

Effective diff

Exactly three files relative to #1002:

• python/fast_mlsirm/report.py — two-line exact-type admission plus stable error;
• tests/test_report_title_callback_safety.py — hostile callback, default-title, and HTML-escaping regressions;
• docs/changelog.d/999-report-title-callback-safety.md — authoritative release fragment.

Unrelated report-limit formatting churn was removed. .Jules/palette.md matches protected main exactly and is absent from the effective diff.

Ownership boundary

No report metric, fit statistic, model-selection result, likelihood, optimizer, or psychometric arithmetic changes. Python only validates the public title scalar before rendering. Keep Draft until fresh exact-head CI/security/review evidence for 528d980... is terminal; predecessor evidence does not transfer.

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: 801dbbdd-4221-4f86-af5d-bd9f707ef3b5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[seonghobae]

@opencode-agent Implement the GREEN on the current exact head 18137218ad3eb825a5b792cb3dd1cf8f568b1b62 of this existing branch.

Required minimal delta:

1. In python/fast_mlsirm/report.py, replace resolved_title = title or _default_title(report_type) with one package-owned title-admission helper that:
   • returns _default_title(report_type) for title is None or an empty exact built-in string;
   • returns a non-empty exact built-in string unchanged;
   • rejects every other value, including str subclasses, with ValueError("title must be a string or None") before truth-value testing, .replace(), or HTML escaping can dispatch caller callbacks.
2. Keep the RED tests in tests/test_report_title_callback_safety.py; do not weaken them.
3. Add one bounded authoritative fragment docs/changelog.d/999-generic-report-title-boundary.md.
4. Do not modify 🎨 Palette: [UX improvement] Improve focus styles for main content #978's focus CSS/tests, report metrics/tables, statistical code, aggregate CHANGELOG.md, locks, workflows, or unrelated files.
5. Run the focused title test and relevant report tests. Commit to this same branch without force-push.

If the branch head moved, refetch it, inspect the delta, and apply only if compatible.

[seonghobae]

@opencode-agent Clean the current exact head 8cddc840aadf4c536f0aa6edb1a58f07cccf1b55 without changing behavior.

In python/fast_mlsirm/report.py, restore protected-main's original multiline formatting for exactly these four pre-existing raise ValueError(...) statements:

• MAX_REPORT_METRICS in _metric_section;
• MAX_REPORT_ROWS and MAX_REPORT_CELLS in _rows_from_columnar;
• MAX_REPORT_ROWS in _validate_report_rows.

Preserve the two intentional changes in the same file:

• main:focus:not(:focus-visible) from former 🎨 Palette: [UX improvement] Improve focus styles for main content #978;
• exact-built-in title admission from fix(report): harden generic diagnostics title trust boundary #999.

Do not touch tests, changelog fragments, CSS beyond that selector, locks, aggregate CHANGELOG, workflows, or other files. Commit to the same branch without force-push after refetching the exact head.

[seonghobae]

@opencode-agent review
@cwl-noema-review review

The draft is ready after exact-head validation. Re-review head 528d980c1613fd6326f4c5c59effc85e80fb98b2 against base fix/report-main-focus-current-review-978; use current source and Checks only. Local generic diagnostics-report title validation passed 3 tests covering hostile title rejection before callbacks/artifact creation, default-title compatibility, and HTML escaping.