oh-my-customcode 1.1.90 → 1.1.92

package/templates/.claude/hooks/hooks.json

@@ -7,7 +7,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/stage-blocker.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/stage-blocker.sh\""
           }
         ],
         "description": "Block Write/Edit tools during plan/verify/compound stages — only allow in implement stage"
@@ -67,7 +67,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/content-hash-validator.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/content-hash-validator.sh\""
           }
         ],
         "description": "Validate file content hash before Edit — advisory staleness warning"
@@ -77,7 +77,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/schema-validator.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/schema-validator.sh\""
           }
         ],
         "description": "Schema-based tool input validation — Phase 1 advisory only"
@@ -87,7 +87,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/rtk-intercept.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/rtk-intercept.sh\""
           }
         ],
         "description": "RTK auto-intercept — transparently rewrites CLI commands through RTK proxy when available (R013 advisory)"
@@ -97,7 +97,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/destructive-git-guard.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/destructive-git-guard.sh\""
           }
         ],
         "description": "Warn before destructive git commands — advisory recovery guidance (R001/R021)"
@@ -107,7 +107,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/rule-deletion-guard.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/rule-deletion-guard.sh\""
           }
         ],
         "description": "Block rule file deletion — requires individual user confirmation per rule (R001 safety)"
@@ -121,15 +121,15 @@
           },
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/git-delegation-guard.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/git-delegation-guard.sh\""
           },
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/agent-teams-advisor.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/agent-teams-advisor.sh\""
           },
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/model-escalation-advisor.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/model-escalation-advisor.sh\""
           }
         ],
         "description": "HUD statusline + R010 git delegation guard + R018 Agent Teams advisor on Agent/Task spawn"
@@ -141,7 +141,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": ".claude/hooks/scripts/omcustom-auto-update.sh"
+            "command": "\"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/omcustom-auto-update.sh\""
           }
         ],
         "description": "Interactive oh-my-customcode update check at session start (#752)"
@@ -151,7 +151,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/session-env-check.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/session-env-check.sh\""
           }
         ],
         "description": "Check codex CLI and Agent Teams availability at session start"
@@ -161,7 +161,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/stale-todo-scanner.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/stale-todo-scanner.sh\""
           }
         ],
         "description": "Scan TODO.md files for stale items at session start"
@@ -171,7 +171,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/adaptive-harness-scan.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/adaptive-harness-scan.sh\""
           }
         ],
         "description": "Lightweight project profile staleness check for adaptive harness (#831)"
@@ -181,7 +181,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/session-autofix.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/session-autofix.sh\""
           }
         ],
         "description": "Auto-detect and fix previous session issues at start (#838)"
@@ -191,7 +191,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/plugin-cache-check.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/plugin-cache-check.sh\""
           }
         ],
         "description": "Advisory check for plugin cache directories missing node_modules (#1207)"
@@ -201,7 +201,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/claude-md-reinject.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/claude-md-reinject.sh\""
           }
         ],
         "description": "Re-inject project CLAUDE.md into model context on compact re-entry only — CC already natively auto-loads CLAUDE.md on startup/resume/clear, so re-injecting there is pure duplication (~13.8KB/session); matcher narrowed to \"compact\" to keep the original #1617 purpose (post-compact amnesia defense) without the startup double-injection (#1626)"
@@ -213,7 +213,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/user-prompt-preprocessor.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/user-prompt-preprocessor.sh\""
           }
         ],
         "description": "Advisory pre-processing of user input — skill matching hints and intent analysis"
@@ -223,7 +223,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/r007-r008-drift-advisor.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/r007-r008-drift-advisor.sh\""
           }
         ],
         "description": "Proactive R007/R008 drift advisory — checks last assistant turn for identification omission (#1229)"
@@ -233,7 +233,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/session-autofix-prompt.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/session-autofix-prompt.sh\""
           }
         ],
         "description": "Inject session auto-fix findings into first user prompt (#838)"
@@ -243,7 +243,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/fail-axis-cause-advisor.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/fail-axis-cause-advisor.sh\""
           }
         ],
         "description": "Advisory on cause-free progress nudges when this session has logged tool failures — delivers via hookSpecificOutput.additionalContext (#1561)"
@@ -259,7 +259,7 @@
           },
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/agent-start-recorder.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/agent-start-recorder.sh\""
           }
         ],
         "description": "HUD display + agent start time recording for stall detection (R009)"
@@ -271,23 +271,23 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/task-outcome-recorder.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/task-outcome-recorder.sh\""
           },
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/stall-detection-advisor.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/stall-detection-advisor.sh\""
           },
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/auto-continue-guard.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/auto-continue-guard.sh\""
           },
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/r007-r008-drift-advisor.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/r007-r008-drift-advisor.sh\""
           },
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/subagent-failure-advisor.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/subagent-failure-advisor.sh\""
           }
         ],
         "description": "Record agent outcomes + auto-continue workflow + R007/R008 drift advisory + failure advisory on subagent completion (autonomous-loop re-entry, #1545; #1631 replaced the self-referential prompt-type judge hook with an advisory-only command hook)"
@@ -299,7 +299,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/cwd-change-detector.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/cwd-change-detector.sh\""
           }
         ],
         "description": "Auto-detect project type on working directory change (v2.1.83+)"
@@ -311,7 +311,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/file-change-validator.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/file-change-validator.sh\""
           }
         ],
         "description": "Validate externally modified files — advisory staleness warning (v2.1.83+)"
@@ -371,7 +371,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/task-state-precompact.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/task-state-precompact.sh\""
           }
         ],
         "description": "Serialize active task state before compaction — pairs with PostCompact restore (#922)"
@@ -469,7 +469,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/context-budget-advisor.sh",
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/context-budget-advisor.sh\"",
             "continueOnBlock": true
           }
         ],
@@ -480,7 +480,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/r007-r008-drift-advisor.sh",
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/r007-r008-drift-advisor.sh\"",
             "continueOnBlock": true
           }
         ],
@@ -491,7 +491,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/stuck-detector.sh",
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/stuck-detector.sh\"",
             "continueOnBlock": true
           }
         ],
@@ -502,7 +502,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/cost-cap-advisor.sh",
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/cost-cap-advisor.sh\"",
             "continueOnBlock": true
           }
         ],
@@ -513,7 +513,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/content-hash-validator.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/content-hash-validator.sh\""
           }
         ],
         "description": "Store content hashes for Read operations — enables Edit staleness detection"
@@ -523,7 +523,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/secret-filter.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/secret-filter.sh\""
           }
         ],
         "description": "Detect potential secrets in Bash/Read/Grep output — advisory warning only"
@@ -533,7 +533,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/audit-log.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/audit-log.sh\""
           }
         ],
         "description": "Append-only audit log for state-changing tool operations"
@@ -543,7 +543,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/skill-count-reminder.sh",
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/skill-count-reminder.sh\"",
             "continueOnBlock": true
           }
         ],
@@ -554,7 +554,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/playwright-compress.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/playwright-compress.sh\""
           }
         ],
         "description": "Layer 4: Compress Playwright/Chrome MCP output via Haiku summarization"
@@ -564,7 +564,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/auto-dev-token-tracker.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/auto-dev-token-tracker.sh\""
           }
         ],
         "description": "auto-dev token spend tracker — records per-phase token estimates during /pipeline auto-dev (Issue #1057, advisory)"
@@ -576,7 +576,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/stop-console-audit.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/stop-console-audit.sh\""
           }
         ],
         "description": "Final console.log audit and session diagnostics before session ends"
@@ -586,7 +586,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/eval-core-batch-save.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/eval-core-batch-save.sh\""
           }
         ],
         "description": "Batch-save agent outcomes to eval-core DB on session end (advisory, exit 0)"
@@ -596,7 +596,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/feedback-collector.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/feedback-collector.sh\""
           }
         ],
         "description": "Auto-extract failure patterns from session outcomes (advisory, exit 0)"
@@ -606,7 +606,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/session-reflection.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/session-reflection.sh\""
           }
         ],
         "description": "Transcript-based self-reflection: detect R007/R008 violations and log to .claude/outputs/reflections/ (Phase 1 MVP, advisory, exit 0)"
@@ -616,7 +616,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/skill-extractor-analyzer.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/skill-extractor-analyzer.sh\""
           }
         ],
         "description": "Detect reusable skill candidates from session task outcomes (advisory, exit 0)"
@@ -626,7 +626,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/auto-dev-token-summary.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/auto-dev-token-summary.sh\""
           }
         ],
         "description": "Print auto-dev token spend summary on session end (Issue #1057, advisory)"
@@ -638,7 +638,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "bash .claude/hooks/scripts/failure-ledger.sh"
+            "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/failure-ledger.sh\""
           }
         ],
         "description": "FAIL-axis instrumentation — append tool failures to the error ledger (JSONL). Never blocks; feeds fail-axis-cause-advisor.sh (#1561)"

package/templates/.claude/hooks/scripts/rule-deletion-guard.sh

@@ -13,7 +13,7 @@ if ! command -v jq &>/dev/null; then
 fi
 
 # Parse tool input
-tool=$(printf '%s\n' "$input" | jq -r '.tool // ""' 2>/dev/null) || { printf '%s\n' "$input"; exit 0; }
+tool=$(printf '%s\n' "$input" | jq -r '.tool_name // .tool // ""' 2>/dev/null) || { printf '%s\n' "$input"; exit 0; }
 cmd=$(printf '%s\n' "$input" | jq -r '.tool_input.command // ""' 2>/dev/null) || { printf '%s\n' "$input"; exit 0; }
 
 # Only check Bash tool

package/templates/.claude/hooks/scripts/schema-validator.sh

@@ -21,7 +21,11 @@ tool_name=$(printf '%s\n' "$input" | jq -r '.tool_name? // "unknown"')
 tool_input=$(printf '%s\n' "$input" | jq -c 'if (.tool_input? | type) == "object" then .tool_input else {} end' 2>/dev/null) || tool_input='{}'
 [ -n "$tool_input" ] || tool_input='{}'
 
-SCHEMA_FILE=".claude/schemas/tool-inputs.json"
+# Locate the repo root ourselves (#1767): hooks run in the session cwd, which may be
+# a subdirectory. Prefer CLAUDE_PROJECT_DIR; fall back to this script's own location
+# (<root>/.claude/hooks/scripts/ -> three levels up).
+ROOT="${CLAUDE_PROJECT_DIR:-$(cd "$(dirname "${BASH_SOURCE[0]}")/../../.." && pwd)}"
+SCHEMA_FILE="$ROOT/.claude/schemas/tool-inputs.json"
 
 # Skip if schema file doesn't exist
 if [ ! -f "$SCHEMA_FILE" ]; then

package/templates/.claude/hooks/skill-count-reminder.sh

@@ -5,6 +5,9 @@
 
 set -euo pipefail
 
+# Locate the repo root ourselves: hooks run in the session cwd, which may be a subdirectory
+ROOT="${CLAUDE_PROJECT_DIR:-$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)}"
+
 # Read tool input from stdin
 INPUT=$(cat)
 
@@ -14,7 +17,8 @@ FILE_PATH=$(echo "$INPUT" | jq -r '.tool_input.file_path // .tool_input.path //
 # Only trigger for .claude/skills/ paths with SKILL.md
 if [[ "$FILE_PATH" == *".claude/skills/"*"SKILL.md"* ]]; then
     SKILL_NAME=$(echo "$FILE_PATH" | sed 's|.*\.claude/skills/||' | sed 's|/SKILL.md||')
-    ACTUAL_COUNT=$(find .claude/skills -name 'SKILL.md' 2>/dev/null | wc -l | tr -d ' ')
+    # '|| true' keeps the advisory contract (exit 0) even if the skills directory is missing under pipefail
+    ACTUAL_COUNT=$( { find "$ROOT/.claude/skills" -name 'SKILL.md' 2>/dev/null || true; } | wc -l | tr -d ' ')
 
     cat >&2 << EOF
 ─── [Skill Sync Reminder] New/modified skill: ${SKILL_NAME} ───

package/templates/.claude/rules/MAY-optimization.md

@@ -78,6 +78,8 @@ zsh는 미인용 `$var`를 단어분할하지 않습니다(#1683 #5) — 분할
 
 계수/매칭 방법 확인(#1521): 카운트 대조 전 비교 대상의 계수 방법(glob vs `find`, 부분매칭, 확장자 필터, 머지커밋 diff 생략 등)을 먼저 확인합니다.
 
+`git grep -E`(POSIX ERE)는 `\b`·`\d`·`\s`를 지원하지 않으며 오류 없이 0건을 반환할 수 있습니다(#1760, 플랫폼 의존 — macOS git 2.52.0에서 확인, GNU/glibc ERE는 확장으로 지원) — 이식 가능한 PCRE가 필요하면 `git grep -P` 또는 python을 사용합니다.
+
 <!-- DETAIL: counting method check
 > **계수/매칭 방법 확인 (#1521)**: 카운트를 대조하기 전에 **비교 대상이 무엇을 어떻게 세는지** 먼저 확인한다 — 같은 지표라도 계수 방법이 다르면 값이 달라진다. 대표 함정 4종: (a) glob(`ls *.md`, 최상위만) vs 재귀 `find`(하위 디렉토리 포함), (b) 부분 문자열 grep(`grep "sdd"`가 `sdd-dev`까지 매칭), (c) 확장자 필터(`--include='*.md'`가 `CLAUDE.md.en`을 미매칭), (d) **머지 커밋 diff 기본 생략** — `git show --name-only <머지커밋>`은 diff를 기본적으로 출력하지 않아 변경 파일 0개로 오독된다. 머지 커밋의 변경 파일을 세려면 `--first-parent`(1차 부모 대비) 또는 `-m`(각 부모별 diff)을 명시한다. 검증 스크립트와 대조할 때는 **스크립트의 실제 계수 로직을 읽고** 같은 방법으로 센다. 위 `ls | tail` 시계열 오판(#1417)과 동류로, 도구의 기본 동작을 확인하지 않은 채 결과를 해석해 오탐에 이르는 패턴이다. Origin: #1521 (2026-07-20 세션에서 3회 반복; 두 서브에이전트가 독립적으로 동일 오탐에 도달); (d)는 #1553 찐빠 #4 (2026-07-30 세션에서 머지 커밋 `--name-only` 0파일을 "변경 없음"으로 오독).
 -->

package/templates/.claude/rules/MUST-enforcement-policy.md

@@ -33,7 +33,7 @@ oh-my-customcode uses an **advisory-first enforcement model**. Most rules are en
 
 | Anti-pattern | Required |
 |--------------|----------|
-| `hooks.json`만 편집하고 커밋 → settings.json `hooks` 블록 미재생성 | 편집 후 `hooks-settings.ts` 변환기(빌드) 실행 → settings.json(+local+templates) 재생성 확인 후 커밋 |
+| `hooks.json`만 편집하고 커밋 → settings.json `hooks` 블록 미재생성 | 편집 후 `bun run sync:hooks`(또는 `bun run build`) 실행 → settings.json(+local+templates) 재생성 확인 후 커밋 |
 
 교훈: **배선 확인 ≠ 전달 확인 ≠ 발화 확인 ≠ 로드 확인** — R020 "actual outcome ≠ attempt"의 훅 도메인 재현 사례. 상세는 Read 도구로 열람.
 

package/templates/.claude/rules/MUST-sync-verification.md

@@ -503,6 +503,8 @@ Origin: #1584 #1 (v1.1.46 세션) — 이슈 생성(8/11~14)과 작업(8/15) 사
 
 R017 게이트 통과 후 스코프가 확장되면(추가 파일 편집), 커밋 전 게이트를 **최종 상태에서 재실행**한다.
 
+**경계 (#1760)**: 이 경계는 위 재실행 규칙을 취소하지 않고 같은 파일 내 리뷰 정정에 한해 좁힙니다. "리뷰 정정"은 이 게이트의 리뷰 위임(mgr-sauron R017, 적대적 리뷰)의 지적에 따라 게이트 대상 변경 집합에 이미 있는 파일만 바꾼 정정입니다. 그런 정정 후 결정론 검사(`bun run lint`, `bun run typecheck`, `bash .github/scripts/verify-template-sync.sh`, `bun .github/scripts/validate-docs.ts --programmatic-only`, 전체 `bun test`)를 최종 상태에서 재통과했다면 구조 재검증(mgr-sauron 재실행)을 생략할 수 있으며 생략 사실을 기록합니다. 이 생략은 게이트 판정이 PASS였던 경우에만 적용되며, FAIL 판정 후의 정정은 PASS가 나올 때까지 항상 게이트를 재실행합니다. #1688이 요구하는 리뷰 후 위키 재동기화가 만든 위키 페이지·매니페스트는 `verify-wiki-sync.sh`가 다루므로 그것만으로는 재실행 사유가 아닙니다. 그 외 새 파일이나 새 구조 표면이 생기면 게이트를 재실행합니다.
+
 <!-- DETAIL: Post-Gate Scope-Expansion Re-Run intro, original wording
 R017 게이트(mgr-sauron) 통과 선언 후 신규 결함 발견 등으로 스코프가 확장되면(추가 파일 편집), 커밋 전 게이트를 **최종 상태에서 재실행**한다. 게이트 통과 시점 이후의 변경은 형식적으로 미검증이므로, 확장분 미검증 커밋은 R017이 최종 산출물을 커버하지 못하게 만든다.
 -->