Skip to content

Add Ohio State University mirror (port 40020) with deterministic verifiers - #63

Merged
Raibows merged 4 commits into
mainfrom
osu-mirror-verifiers
Sep 7, 2026
Merged

Add Ohio State University mirror (port 40020) with deterministic verifiers#63
Raibows merged 4 commits into
mainfrom
osu-mirror-verifiers

Conversation

@StephenQSstarThomas

@StephenQSstarThomas StephenQSstarThomas commented Jul 15, 2026

Copy link
Copy Markdown
Contributor

Add Ohio State University mirror as site 21 on port 40020

This branch now includes the original OSU mirror and grading contract, current main through TED, and the review remediation for PR #63.

Integration

  • OSU is appended after TED as site index 20 and runs on container port 40020.
  • websyn_start.sh, control_server.py, Dockerfile, contributor documentation, and task URLs use the same 21-site order and 40000-40020 range.
  • sites/osu/.build-generated-seed marks its deterministic source-generated database. The pinned Hugging Face asset revision contains the required osu.tar.gz official-image bundle, enforced by sites/osu/.requires-images, scripts/build.sh, scripts/check_assets.sh, and the Docker build.

Review remediation

  • Removed the committed Flask secret, open redirects, GET logout, and news GET mutation.
  • Added request/session protections, SQLite foreign keys, bookmark type/object validation, a uniqueness constraint, race handling, note limits, and robust login/registration behavior.
  • Made seed initialization module-safe, partial-state aware, and deterministic; fixed event filtering to use the benchmark clock.
  • Added token-overlap ranked site search and corrected mobile/responsive, focus, navigation, and accessibility behavior.
  • Crawled and normalized 19 photographs from official Ohio State web properties. sites/osu/image_sources.json records source pages, source URLs, dimensions, alt text, and source/output SHA-256 values; sites/osu/fetch_images.py reproduces the WebP bundle.
  • Reworked all 20 tasks into explicit visible workflows with multi-field facts, filters, and comparisons.
  • Replaced all 20 verifiers with same-origin exact path/query/order checks, visible-link transition checks, negation/entity binding, and complete before/after database equality.

Validation

  • 24 OSU application, seed, image-provenance, integration, and verifier tests pass.
  • 20/20 actual Playwright task runs and their deterministic verifiers pass.
  • 45/45 responsive checks at 320 px, 390 px, and 1440 px pass with no page-level horizontal overflow.
  • Hugging Face dataset PR Add CarMax task verifiers + judge rubrics #56 is merged; pinned revision db9c73e62d853ed91f6152b5b7105571502b7e07 contains osu.tar.gz.
  • Clean Docker build succeeds with no host OSU seed present and requires the official-image bundle.
  • All 21 sites are healthy and return HTTP 200; /reset/osu and /reset-all pass.

The consolidated audit is in review-reports/PR-63-FINAL-AUDIT.md.

StephenQSstarThomas and others added 3 commits July 15, 2026 05:04
Integrate the osu mirror from PR #12 onto the current main as an additive 17th
site instead of the PR's stale base (which replaced merriam_webster and reused
port 40015). Registered in all three places (websyn_start.sh, control_server.py,
Dockerfile EXPOSE 40000-40016); the readiness gate in websyn_start.sh is now
derived from ${#SITES[@]} so it no longer hardcodes the site count.

osu ships no frozen DB/images in git or HF: its data is generated from
seed_data.py, so a Dockerfile build step creates instance_seed/osu.db so the
boot reset and /reset/osu have a seed to restore from. Verified: image builds,
osu serves on 40016, and /reset byte-identity holds.
Reviewer role for PR #12 (previously unreviewed). One verify_N.py per task over a
shared verify_lib.py, mirroring sites/merriam_webster/verify: deterministic-first
(navigation anti-shortcut + hardcoded on-site ground truth) with an anchored
llm_screenshot_shows confirmation that is skipped under --no_llm. tasks.jsonl gains
verifier_path + judge_rubric per row (ground truth stays out of the agent-facing
file). selfcheck.py proves all 20 verifiers PASS a correct trajectory and FAIL a
wrong-answer / no-navigation one (20/20). TASK_REVIEW.md records feasibility: all
20 tasks are on-site answerable; 6 are flagged as answerable-from-prior-knowledge
(mitigated by the nav check) and task 1 has a 36-vs-27 varsity-sports inconsistency.
@Raibows Raibows changed the title Add Ohio State University mirror site (port 40016) + deterministic task verifiers Add Ohio State University mirror (port 40020) with deterministic verifiers Sep 7, 2026
@Raibows

Raibows commented Sep 7, 2026

Copy link
Copy Markdown
Contributor

Thanks for your contribution!

@Raibows
Raibows merged commit 7269134 into main Sep 7, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants