fix(parity): drive semseg goldens with the CLI's mask_id schema (di#358)

[LukasWodka]

Summary

Fixes the red "Validator goldens vs data-ingestors @ pin" check that appeared on develop after cli#286 bumped the pin past data-ingestors #358.

Root cause — a harness bug, not the CLI or the validator

di#358's MaskIdColumnValidator requires mask_id to be a declared schema column (an undeclared mask_id is dropped at ingest → stored table lacks it → the training client raises FileNotFoundError, backend#816). The CLI's real semseg ingest does declare it (spec.go: schema={"mask_id":"VARCHAR(255)"}) — but gen-validator-goldens.py only injected a schema for tabular/time-series cases, so it drove the validators without the schema the CLI sends and spuriously rejected every semseg case (including semseg-ok).

Fix

1. gen-validator-goldens.py: mirror spec.go — inject {"mask_id":"VARCHAR(255)"} for semantic_segmentation cases (an explicit per-case schema still wins, so reject-path cases keep exercising the contract).
2. semseg-missing-mask-id: flip ingestor_verdict accept→reject — it has no mask_id column, so post-di#358 the ingestor correctly rejects it. This is exactly the reconciliation the case's own note predicted at pin-bump time; CLI and ingestor now agree.

Verification

• sync-validator-goldens.sh --check → validator goldens in sync.
• Goldens diff is one entry (semseg-missing-mask-id: accept→reject, with the real reject message now populated). semseg-ok was already accept — the harness just stops falsely flagging it.
• go test ./internal/push/... green (incl. the cli_verdict ↔ ingestor_verdict ↔ goldens cross-check). Ran locally against the pinned data-ingestors (semseg validator identical between the pin and current DI HEAD).

cc @divyasinghds — this is the di#358 (backend#816) contract you enforced in the validator; flagging in case you want to eyeball the reconciliation.

🤖 Generated with Claude Code

---

Note

Low Risk
Test harness and generated parity fixtures only; no production ingest or CLI preflight logic changes in this diff.

Overview
Aligns the validator-parity harness with how the CLI actually ingests semantic segmentation: after the data-ingestors pin passed di#358, goldens were out of sync because the Python golden generator did not pass the same mask_id schema the CLI sends.

scripts/gen-validator-goldens.py now sets schema / full_schema to {"mask_id": "VARCHAR(255)"} for semantic_segmentation cases (matching spec.go), while an explicit per-case schema still overrides—so happy paths like semseg-ok are no longer spuriously rejected in goldens.

Parity fixtures update semseg-missing-mask-id: ingestor_verdict flips from accept to reject, with an updated note that CLI and pinned ingestor now both enforce the MaskIdColumnValidator contract. goldens.json reflects the real reject message for that case.

^{Reviewed by Cursor Bugbot for commit cb7134d. Bugbot is set up for automated code reviews on this repo. Configure here.}

[LukasWodka]

@BugBot run

[cursor]

✅ Bugbot reviewed your changes and found no new issues!

Comment @cursor review or bugbot run to trigger another review on this PR

^{Reviewed by Cursor Bugbot for commit cb7134d. Configure here.}

[LukasWodka]

👋 Heads-up — Code review queue is at 49 / 30

Above the WIP limit. The team convention is to review existing PRs before opening new work.

Open PRs currently in Code review (oldest first):

• averaging-service#182 — feat(weights): averaging writes SafeTensors (TF + PyTorch) — SEC-03 §8 step 4 · author: @shujaatTracebloc · reviewer: @divyasinghds
• averaging-service#183 — docs: SafeTensors deployment-order & resume-safety runbook (epic feat(auth): logout revokes the token server-side (cli#112) #116) · author: @shujaatTracebloc · reviewer: @saadqbal
• averaging-service#184 — test(weights): shared SafeTensors golden-corpus drift guard + declare safetensors dep (SEC-03) · author: @LukasWodka · no reviewer assigned
• averaging-service#185 — test(weights): average a real SafeTensors edge + a legacy edge in one cycle — SEC-03 §8 · author: @LukasWodka · no reviewer assigned
• backend#1093 — chore(deps): bump django from 5.2.14 to 5.2.15 · author: @dependabot · no reviewer assigned
• backend#1108 — test(api): CLI response-shape contract fixtures for the 9 CLI endpoints (WS-D.2) · author: @LukasWodka · reviewer: @saadqbal
• backend#1111 — feat(download): edge weights fetch tolerates .safetensors (SEC-03 step 5c hardening) · author: @shujaatTracebloc · reviewer: @saadqbal
• backend#1112 — fix(runners): migrate deprecated trainingClasses/getTrainingPlan to snake_case (#1102) · author: @LukasWodka · reviewer: @saqlainsyed007
• backend#1114 — perf(leaderboard): serializer prefetch reads + TeamSerializer is_* in Python (#978 pt2) · author: @aptracebloc · reviewer: @saqlainsyed007
• cli#302 — ci: re-enable staticcheck (pinned standalone) + fix the 5 non-ST1005 findings · author: @LukasWodka · no reviewer assigned

Pull from review before opening new work. (This is a nudge from the kanban WIP check, not a block.)