Engineering Quality Gate — focused policy

Policy for using the Engineering Quality Gate as a focused architecture plus standards-backed implementation review gate.

Scope and review order

Use this policy when the review scope is intentionally limited to architecture boundaries and standards-backed implementation correctness.

  • This policy routes the review to Architecture Boundary Review, Standards-Backed Implementation Review, or both in sequence.
  • Use the full code-review policy when the task includes test coverage, regression risk, refactor planning, general code quality, or code generation.
  • If both focused reviews are required, run Architecture Boundary Review first, then Standards-Backed Implementation Review.

Policy rules

These rules define the focused gate before a concrete review procedure runs.

  1. The gate is valid only when the task is about architecture boundaries, layering, ownership, dependency direction, standards-backed implementation correctness, or both.
  2. Choose Architecture Boundary Review when the dominant issue is structural. Choose Standards-Backed Implementation Review when the dominant issue is correctness against authoritative guidance.
  3. Architecture findings and standards-backed implementation findings must remain separate. If both reviews are needed, run them as two passes.
  4. Architecture review requires structural evidence. Standards-Backed Implementation Review requires code or configuration plus authoritative guidance or a clearly supplied source boundary.

Choose the active review policy

The Engineering Quality Gate policy routes the review. The selected child policy controls the concrete review run.

Use a different policy when the scope is broader

This focused gate must not absorb adjacent engineering workflows.

Verification checklist

Use this checklist before accepting that the focused gate was applied correctly.

  • The run did not expand into testing, regression risk, refactor planning, general code quality, or code generation.
  • Architecture review was used for structure and boundaries. Standards-backed implementation review was used for authoritative-guidance correctness.
  • Architecture findings were supported by structural evidence, and implementation findings were supported by code/configuration plus authoritative guidance or supplied source material.
  • Architecture findings, implementation findings, missing evidence, and next actions were not merged into one undifferentiated review output.