yalla-reviewlisted

# /yalla-review Binary pass/fail code review. Each reviewer checks one specific aspect. No severity scales — either the code passes the check or it doesn't. The default comparison base is `$BASE_BRANCH` (from `base_branch:` in `.claude/YALLA.md`, default `main`). ## Core Principle Each reviewer answers ONE specific question with Pass or Fail. Not "rate the quality" — that produces unactionable noise. Binary forces explicit definitions of what constitutes failure, making disagreements resolvable and fixes obvious. Risk gates are triggered by the changed files and workflow. Do not run every historical check on every PR. The universal checks stay small; subsystem checks run only when their trigger applies. For non-trivial work, include the operator-understanding layer (see `${CLAUDE_PLUGIN_ROOT}/knowledge/yalla/`). This check verifies that the operator/maintainer can understand the change without reading code; it does not quiz the operator or grade writing polish. --- ## Step 1: Get Changed Files ```bash git diff "$BASE_BRANCH" --name-only git diff "$BASE_BRANCH" ``` Also read `${CLAUDE_PLUGIN_ROOT}/knowledge/yalla/ARTIFACTS.md`, `${CLAUDE_PLUGIN_ROOT}/knowledge/yalla/TEST-SEAMS.md`, `${CLAUDE_PLUGIN_ROOT}/knowledge/yalla/ARCHITECTURE-DEPTH.md`, `${CLAUDE_PLUGIN_ROOT}/knowledge/yalla/PROJECT-CHECKS.md`, `${CLAUDE_PLUGIN_ROOT}/knowledge/yalla/REVIEW-CHECKS.md`, and `${CLAUDE_PLUGIN_ROOT}/knowledge/product/INTENDED-VS-IMPLEMENTED.md` before launching reviewers. ## Step 2