Repository navigation
fix(IPO-013): track current evaluations and semantic freshness - #150
Open
DoRmAmMu1997 wants to merge 8 commits into
Open
DoRmAmMu1997 wants to merge 8 commits into
DoRmAmMu1997 wants to merge 8 commits into
Conversation
…checks Separate immutable evaluation content from current selection; guard publication and input writes, bound conflicts, and expose historical stale scores without actionable recommendations. Co-authored-by: Codex <codex@openai.com>
Preserve merged security and dependency changes and current Python matrix before fresh verification. Co-authored-by: Codex <codex@openai.com>
Reject legacy scorecard collisions without mutating receipts or selection; retain semantic source-set equivalence and complete snapshot error contracts. Co-authored-by: Codex <codex@openai.com>
Preserve the merged BSE transcript and agent-result compatibility fix fromPR147. Co-authored-by: Codex <codex@openai.com>
Keep the historical lookup exports while recording the added freshness API. Co-authored-by: Codex <codex@openai.com>
DoRmAmMu1997
marked this pull request as ready for review
October 4, 2026 17:19
Preserve the reviewed feature while aligning current SDK, database, typing and lint pins. Co-authored-by: Codex <codex@openai.com>
Preserve the reviewed IPO current-state contract while adopting the six reviewed dependency updates from main. Co-authored-by: Codex <codex@openai.com>
Reconcile the reviewed package with main 9ded279 and its current dependency pins. Fresh verification follows this merge. Co-authored-by: Codex <codex@openai.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
IPO dashboards previously selected the newest historical evaluation even when evidence reverted to an earlier input set or became stale with time. This adds a separate current-selection record, transactional input revisions, guarded publication with at most three attempts, and fingerprint/model/time freshness checks. A → B → A selects the existing A receipt while keeping original calculation history immutable; stale verdicts remain visible as history and are excluded from actionable recommendations.
The migration
20260921ipo013follows20260909valid005and backfills the newest complete historical pair as unverified. Source registration, manual evidence, demand and enrichment writers participate in the revision boundary. The dashboard rechecks current state on rerun, and job outcomes expose retryable contention explicitly.Independent review found that a legacy caller could seed an arbitrary scorecard using the genuine fingerprint. The follow-up compares the complete persisted score/recommendation payload with the newly derived candidate before selecting an existing pair. A mismatch fails without rewriting history or updating selection. Regressions cover that collision, every receipt component, legitimate source-set reordering, A → B → A, clock expiry, input races and migration history. Google-style contracts and Beginner notes explain the authority, concurrency and retry rules.
Validation on head
be83ae7a, merged with current main9ded279and its updated pins: 2,356 passed, two skipped; 90.68% coverage including app.py on Python 3.12, using twelve commit-locked restartable groups. Pre-commit configuration, compileall, Ruff, full mypy, Bandit, pinned dependency audit and diff checks passed. Python 3.14 passed all 116 focused policy/scoring/dashboard/model/migration cases. Python 3.13 passed 115 initially; the unchanged 15-second Alembic CLI check timed out under load, then passed alone in 4.42 seconds without changing source, assertions or its deadline. Providers are mocked and databases disposable. All seven hosted checks passed on this exact head: Python 3.12–3.14 quality and Docker and CodeQL.Independent scoped re-review closed both the historical-payload and documentation findings. The PR's existing ready state is preserved. The stacked database-verification PR #155 has passed real PostgreSQL transaction/migration tests on all three supported Pythons and its final-image smoke at head437a364 (run38402347283). The broader modernization plan, proposal approval/editor package and final integrated security review remain in progress. No merge or deployment is included.
Co-authored-by: Codex codex@openai.com