fix(privacy): keep public healthz detail-free

[seonghobae]

Public readiness confidentiality boundary

Tracks Issue #182 on the existing non-default branch. The bounded product behavior remains one public-response projection plus its focused regression; no broad health-listener, deployment, schema, credential, provider, or release contract is introduced.

Exact reconstructed topology

Fresh identities after non-destructive current-main reconstruction:

• protected main: a70866e47b74eb74654aa0f8dd12f72260a5bbe2;
• exact contributor head: 97b223b2bcdbd43ac326af7509aa00fa491fd817;
• the head is a two-parent merge result that preserves the prior feature ancestry and includes exact current protected main as its second parent;
• GitHub comparison reports merge base exactly protected main, 0 behind, with a net delta limited to pg_llm_batch/health.py and tests/test_health_public_boundary_current.py;
• the production delta remains one line: serve_healthz() serializes public_health_report(check_health(dsn)) instead of the raw internal report;
• GitHub reports Draft and mergeable; formal reviews and unresolved inline review threads are zero.

The reconstruction used the exact current-main tree and replaced only the reviewed health blob plus its regression blob. No force-push, destructive rebase, conflict-side selection, protected-main write, or historical evidence transfer was used.

Test-first behavior

The regression injects a credential-shaped PostgreSQL URI into internal readiness detail plus an unknown internal probe and requires HTTP /healthz to expose only the fixed database, pg_tiktoken, and com_config boolean projection.

This preserves detailed check_health() evidence for trusted local/operator callers while preventing DSNs, lower-layer error text, arbitrary component names, and topology detail from becoming public HTTP response data.

Corrected historical writer classification

The retained agent/redacted-healthz-readiness ref belongs to closed, unmerged PR #70. It is not an unowned no-PR writer. #70 carried a much broader listener/timeout/header/deployment/documentation hardening proposal and remains historical provenance; this PR neither claims to supersede every broader #70 capability nor copies its stale source result.

That closed residual ref therefore does not by itself block this bounded current-main privacy repair. Any still-desired broader health-service controls must be reconsidered separately from a fresh protected base rather than smuggled into this one-line disclosure fix.

Current exact-head evidence

The prior 2694256d9b5eed336ad9a7ac3345e387001e3d0c runs are predecessor evidence only and do not transfer.

For exact reconstructed head 97b223b2bcdbd43ac326af7509aa00fa491fd817 at the latest read:

• Release Acceptance 31764029663: completed / success;
• Security Scan 31764029691: completed / success;
• SAST Semgrep 31764029722: in progress;
• CI 31764029902: queued.

Queued or in-progress evidence is not acceptance. Required organization workflows and their actual checkout commits must also complete successfully on this unchanged head.

Promotion boundary

Keep Draft until one unchanged exact reconstructed head satisfies Python 3.10/3.12/3.14, exact 100% owned production statement/branch coverage and public docstrings, lint/package/lock/container/reproducibility, Security Scan, SAST, Release Acceptance, every live required organization workflow, zero valid unresolved findings, live ruleset/protection, and every approval requirement actually imposed by policy.

Do not broaden this PR onto canonical PRD/TRD/threat/operability surfaces while Draft #93 owns those documents. No historical #70, predecessor, synthetic-merge, status-only, queued, in-progress, rate-limited, infrastructure-only, or author-only evidence transfers.

Refs #182.

[coderabbitai]

Warning

Review limit reached

@seonghobae, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 59 minutes

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 9d06d43b-cb0d-467b-93c6-a22756fb3ab8

📥 Commits

Reviewing files that changed from the base of the PR and between a70866e and 97b223b.

📒 Files selected for processing (2)

• pg_llm_batch/health.py
• tests/test_health_public_boundary_current.py

---

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.}