Enable Phase-2 id↔label enforcement (blocking gate)#62
Merged
Conversation
The label-correspondence CI was report-only. Now that recipe id↔label drift is zero (regroundings/de-groundings + canonical organism/env labels + jcm.grmd ignored) and the SSSOM generator emits canonical object labels, flip the gate to blocking: - CI workflow: build the SSSOM product (output/ is gitignored), upload the drift report (non-blocking triage artifact), then run `just validate-products` (Engine B) as a BLOCKING step — fails on any ERROR-class verdict across recipe YAMLs + the SSSOM product. - qc recipe: build SSSOM + validate-products (local Phase-2 gate). - NEXT_TASKS.md item 1 marked done. Engine B (single-process) is the gate. validate-terms-all (per-file Engine A) is deliberately NOT wired in — it reloads schema+OAK per file (~hours over ~16k files) and re-checks the same canonical-label surface Engine B covers in one pass. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The label-correspondence job ran the full corpus+OAK validator TWICE (drift report + enforce), ~doubling the job time on every green run. Run the blocking `validate-products` once; regenerate + upload the drift-report artifact only on failure, where it's actually needed for triage. Co-Authored-By: Claude Fable 5 <noreply@anthropic.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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
Flips the
label-correspondenceCI from report-only (Phase 1) to a blocking gate (Phase 2), now that the id↔label drift backlog is zero.Why now
Prereqs are all merged: recipe ID_NOT_FOUND → 0 (#56), kg_fallback cleaned (#59), organism/env labels canonical + jcm.grmd ignored (#53/#57), and the SSSOM generator emits canonical object labels (#58).
just validate-products(Engine B) was verified locally at exit 0.Changes
label-correspondence.yaml): build the SSSOM product (it's gitignored, so not checked in) → upload the drift report (non-blocking triage artifact) → runjust validate-productsas a blocking step. Fails on any ERROR-class verdict (MISMATCH / ID_NOT_FOUND / EMPTY_LABEL / MISSING_COLUMN / MISSING_GLOB / UNKNOWN_PREFIX) across recipe YAMLs + the SSSOM product.qcrecipe: build SSSOM +validate-products(local Phase-2 gate).Engine A vs B
Engine B (single-process, ~2-3 min) is the gate.
validate-terms-all(per-filelinkml-term-validator, Engine A) is deliberately not wired into CI/qc — it reloads schema+OAK per file (~8 h over ~16k files, measured) and re-checks the same canonical-label surface Engine B already enforces in one pass. It stays a targeted dev tool.This PR's own CI exercises the new blocking gate.
🤖 Generated with Claude Code