Files
ai-contract-platform/project-docs/V2-performance-gate-closeout-2026-07-12.md
T
yuqei a3bb78e06b
Security Quality Gate / npm audit, Bandit, pip-audit, detect-secrets (pull_request) Failing after 16m35s
release: freeze V2 release candidate
2026-07-12 15:32:24 +08:00

2.3 KiB

V2 Performance Gate Closeout - 2026-07-12

Scope

Release-prep performance gate follow-up for the only failing scenario from the previous interface benchmark: contract_detail.

Root Cause

The contract detail read path loaded review_tasks with the contract detail payload, then load_contract_action_context(...) queried latest review state again and always queried approval and obligation context.

For contracts whose latest review result already determines next_action, the approval and obligation queries are not used by the display derivation.

Change

  • Reused already-loaded ContractModel.review_tasks in load_contract_action_context(...).
  • Short-circuited approval and obligation context lookups when review state already takes display priority.
  • Skipped obligation lookup for statuses that cannot display obligation context.
  • Added a regression test proving structured review detail reads query review_tasks once and do not query approval_instances or obligation_tasks.

Local Verification

python -m compileall -q yuqei-contract-platform/services/contract-api/src/yuqei_contract_api
python -m pytest yuqei-contract-platform/services/contract-api/tests/test_contract_workflow.py -k "structured_review_results_drive_next_action or reuses_loaded_review_context or contract_next_action_uses_open_obligation_context or structured_action_context_overrides_action_hint" -q
python -m pytest yuqei-contract-platform/services/contract-api/tests/test_contract_workflow.py -k "next_action or query_read_cache or sqlalchemy_repository_persists" -q

All checks passed.

Smoke Verification

Latest smoke deploy passed:

  • project-docs/qa-artifacts/v2-smoke-docker-deploy/2026-07-12T00-41-17

Latest interface benchmark passed:

  • project-docs/qa-artifacts/v2-interface-benchmark/2026-07-12T00-50-00

Contract detail improved from the prior failed benchmark:

Metric Previous failed run Current passed run
p50 87.82 ms 23.53 ms
p95 2032.09 ms 27.95 ms
p99 2241.75 ms 28.35 ms
errors 0 0

Latest release quality gate passed:

  • project-docs/qa-artifacts/v2-release-quality-gate/2026-07-12T00-47-08

Included checks:

  • Contract API schema-check: pass
  • AI Platform schema-check: pass
  • Docker smoke summary freshness: pass
  • Interface benchmark summary freshness: pass