oh-my-customcode 1.1.93 → 1.1.95

package/templates/.claude/hooks/scripts/session-autofix.sh

@@ -6,6 +6,13 @@
 # Time budget: <3s
 
 input=$(cat)
+
+# Anchor cwd to the project root (#1770): hooks may be launched from a subdirectory.
+# Fallback is the git toplevel (or pwd), NOT the script location: tests run these scripts
+# from a tmp cwd with inherited env. (schema-validator.sh/skill-count-reminder.sh instead
+# use a script-location fallback.) The guarded cd must not abort under set -e (e.g. mode-000 dir).
+ROOT="${CLAUDE_PROJECT_DIR:-$(git rev-parse --show-toplevel 2>/dev/null || pwd)}"
+{ [ -d "$ROOT" ] && cd "$ROOT"; } 2>/dev/null || true
 FIXES_FILE="/tmp/.claude-session-fixes-${PPID}"
 LOG_DIR=".claude/outputs/session-fixes"
 ISSUES=()

package/templates/.claude/hooks/scripts/session-env-check.sh

@@ -9,6 +9,13 @@ set -euo pipefail
 
 input=$(cat)
 
+# Anchor cwd to the project root (#1770): hooks may be launched from a subdirectory.
+# Fallback is the git toplevel (or pwd), NOT the script location: tests run these scripts
+# from a tmp cwd with inherited env. (schema-validator.sh/skill-count-reminder.sh instead
+# use a script-location fallback.) The guarded cd must not abort under set -e (e.g. mode-000 dir).
+ROOT="${CLAUDE_PROJECT_DIR:-$(git rev-parse --show-toplevel 2>/dev/null || pwd)}"
+{ [ -d "$ROOT" ] && cd "$ROOT"; } 2>/dev/null || true
+
 echo "" >&2
 echo "--- [Session Environment Check] ---" >&2
 

package/templates/.claude/hooks/scripts/stale-todo-scanner.sh

@@ -9,6 +9,13 @@ set -euo pipefail
 
 input=$(cat)
 
+# Anchor cwd to the project root (#1770): hooks may be launched from a subdirectory.
+# Fallback is the git toplevel (or pwd), NOT the script location: tests run these scripts
+# from a tmp cwd with inherited env. (schema-validator.sh/skill-count-reminder.sh instead
+# use a script-location fallback.) The guarded cd must not abort under set -e (e.g. mode-000 dir).
+ROOT="${CLAUDE_PROJECT_DIR:-$(git rev-parse --show-toplevel 2>/dev/null || pwd)}"
+{ [ -d "$ROOT" ] && cd "$ROOT"; } 2>/dev/null || true
+
 TODO_FILES=("TODO.md" ".claude/TODO.md")
 NOW=$(date +%s)
 FOUND_ANY=false

package/templates/.claude/rules/SHOULD-hud-statusline.md

@@ -51,11 +51,11 @@ Format: `─── [Spawn] {subagent_type}:{model} | {description} ───`
 
 Format: `{Cost} | {project} | {branch} | RL:{rate_limit}% {countdown} | WL:{weekly_limit}% {countdown} | CTX:{usage}%`
 
-Config in `.claude/settings.local.json`: `statusLine.type: "command"`, `statusLine.command: ".claude/statusline.sh"`. Requires CC v2.1.80+ for RL/WL segments. `refreshInterval` setting (v2.1.97+): Auto-refresh interval in seconds for the status line command. Set in `statusLine.refreshInterval` in settings.json.
+Config in `.claude/settings.local.json`: `statusLine.type: "command"`, `statusLine.command: "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/statusline.sh\""` (anchored form written by the installer; the old bare `.claude/statusline.sh` default is migrated on update/init). Requires CC v2.1.80+ for RL/WL segments. `refreshInterval` setting (v2.1.97+): Auto-refresh interval in seconds for the status line command. Set in `statusLine.refreshInterval` in settings.json.
 
 <!-- DETAIL: Statusline configuration JSON and color coding
 ```json
-{ "statusLine": { "type": "command", "command": ".claude/statusline.sh", "padding": 0 } }
+{ "statusLine": { "type": "command", "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/statusline.sh\"", "padding": 0 } }
 ```
 Color coding: Cost (<$1 green, $1-4.99 yellow, >=5 red), RL/WL (<50% green, 50-79% yellow, >=80% red), CTX (<60% green, 60-79% yellow, >=80% red).
 Countdown format: >=1d → "{d}d{h}h", >=1h → "{h}h{m}m", <1h → "{m}m", unavailable → omitted.

package/templates/.claude/skills/claude-native/SKILL.md

@@ -182,13 +182,13 @@ Semver comparison: major → minor → patch (all numeric). Pre-release suffixes
 
 ### Automatic (SessionStart Hook)
 
-Can be integrated into the SessionStart hook to check for new releases at session start:
+Can be integrated into the SessionStart hook to check for new releases at session start (illustrative example — `claude-native-check.sh` is a hypothetical script you would write yourself; it is not shipped or registered in `hooks.json`):
 
 ```json
 {
   "SessionStart": [
     {
-      "command": "bash .claude/hooks/scripts/claude-native-check.sh"
+      "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/claude-native-check.sh\""
     }
   ]
 }

package/templates/.claude/skills/homework/SKILL.md

@@ -145,6 +145,8 @@ Assemble a feedback issue in Korean using the #1266 format:
 
 If `--severity` filter is active, include only findings at or above the threshold. Note the filter in the issue body.
 
+When a 하네스 제안 item contains a runnable command, run it once at draft time (positive and negative fixture where applicable) and record the observed exit code/output next to that proposal; a command that only reports failure must be caught here, not in review (e.g. `bunx biome check` exits 0 inside biome's `files.includes` scope but 1 outside it, so a norm requiring it for all TS/JS reports failure for out-of-scope files regardless of formatting). (#1781 찐빠 #1)
+
 If no findings are discovered, output:
 ```
 [homework] 이번 세션에서 찐빠를 발견하지 못했습니다. (세션 정상 종료)

package/templates/.claude/skills/pipeline/workflows/auto-dev.yaml

@@ -413,7 +413,7 @@ steps:
 
   - name: deep-plan
     skill: deep-plan
-    description: "Research-validated implementation plan (research → plan → verify) — skipped if docs-only, integrated-analysis allowed if lite; independent of tier, MAY run in lightweight mode when compression-mode-eval's 'Cross-tier — Lightweight Skill-Mode Substitution' conditions are met. MULTI-PHASE: 스킬을 그대로 spawn하지 말고 R020 「위임 경계를 Phase 개수로 설계」에 따라 단일 목표 위임으로 분할해 순차 발주한다 (#1595 #4). 리서치·계측 위임의 완료 조건에는 결론 수치를 산출물 표에서 jq/awk로 재계산해 병기하도록 명시한다(#1707 #4). 테스트 격리·모듈 모킹 수정을 다루는 리서치·계획 위임에도 대상 src 파일과 All files의 수정 전후 커버리지를 동일 하네스로 비교하도록 완료 조건에 명시합니다(#1778 찐빠 #5, implement 스텝 Rules와 동일 요구)."
+    description: "Research-validated implementation plan (research → plan → verify) — skipped if docs-only, integrated-analysis allowed if lite; independent of tier, MAY run in lightweight mode when compression-mode-eval's 'Cross-tier — Lightweight Skill-Mode Substitution' conditions are met. MULTI-PHASE: 스킬을 그대로 spawn하지 말고 R020 「위임 경계를 Phase 개수로 설계」에 따라 단일 목표 위임으로 분할해 순차 발주한다 (#1595 #4). 리서치·계측 위임의 완료 조건에는 결론 수치를 산출물 표에서 jq/awk로 재계산해 병기하도록 명시한다(#1707 #4). 테스트 격리·모듈 모킹 수정을 다루는 리서치·계획 위임에도 대상 src 파일과 All files의 수정 전후 커버리지를 동일 하네스로 비교하도록 완료 조건에 명시합니다(#1778 찐빠 #5, implement 스텝 Rules와 동일 요구). 적용 경계 표에는 각 행마다 이 위치를 읽는 위임 유형/단계 열을 두고 결함이 발생한 단계와 일치하는지 표에서 대조합니다(#1781 찐빠 #2). 기본값·출력 리터럴을 바꾸는 계획은 그 리터럴을 단언하는 기존 테스트를 git grep으로 전수 열거해 웨이브 표의 파일 소유권에 포함합니다(git grep은 untracked 파일을 보지 못하므로 git status --porcelain도 확인합니다, #1786 찐빠 #5)."
     depends_on: plan
 
   - name: implement
@@ -486,6 +486,9 @@ steps:
         its lightweight tier substitute, applying to lite and standard tiers, and
         implement-stage delegation prompts MUST carry the relevant boundary row for each edit
         (회고 제안 묶음 릴리즈의 통합 분석에 적용 경계 표 필수, #1735 찐빠 #4).
+        The table MUST include a column naming which delegation type/step reads each target
+        location and be checked against the step where the defect occurred; a mismatch means
+        the clause must also be placed at the location that step reads (#1781 찐빠 #2).
       - Before dispatching any delegation prompt in an iteration that created or reinforced a rule
         clause, the orchestrator MUST compare that clause's Anti-pattern table rows 1:1 against the
         delegation prompt's sentences and rewrite any match (cause: the R016 「신설 조항의 동일
@@ -502,7 +505,7 @@ steps:
         block: (a) Korean 합쇼체 for new sentences, do not imitate adjacent 반말 — completion
         criteria MUST include THREE deterministic checks, scoped to edited text files (md/yaml)
         and restricted to added lines only (`git diff -U0 -- <edited md/yaml files> | grep
-        '^+'`): family check `grep -cE '(한다|된다|않는다|따른다|했다|있다|없다|넣는다|이다)[.。 "]'`
+        '^+'`): family check `grep -cE '(한다|된다|않는다|따른다|했다|있다|없다|넣는다|이다)[.。 "(]'`
         = 0, line-final auxiliary check `grep -E '다[.。"]?$' | grep -vcE '니다[.。"]?$'` = 0
         (the family list alone misses endings such as 따른다/했다/있다 and unpunctuated
         line-final endings; the line-final check excludes 합쇼체 `-니다` endings so it does not
@@ -521,7 +524,7 @@ steps:
         python3 `re` equivalent of all THREE checks above, reading from the SAME piped stdin as
         (a) (`git diff -U0 -- <edited md/yaml files> | grep '^+' | python3 -c "import re,sys;
         text=sys.stdin.read();
-        fam=len(re.findall(r'(한다|된다|않는다|따른다|했다|있다|없다|넣는다|이다)[.。 \"]', text));
+        fam=len(re.findall(r'(한다|된다|않는다|따른다|했다|있다|없다|넣는다|이다)[.。 \"(]', text));
         lf=len(re.findall(r'다[.。\"]?$', text, re.M)) - len(re.findall(r'니다[.。\"]?$', text, re.M));
         noun=len(re.findall(r'(함|됨|필요)[.。\"]?$', text, re.M));
         print(fam, lf, noun)"`, NOT a file-argument read — the input is the piped diff, never a
@@ -550,7 +553,11 @@ steps:
         delegation creates MUST live under a per-agent-unique path (`$TMPDIR` or the session
         scratchpad), never a fixed shared path (#1722 찐빠 #7). This governs the agent's own
         ad-hoc temp files; temp paths created at runtime by test code are governed by the
-        separate test-code temp-path bullet below (#1778 찐빠 #4).
+        separate test-code temp-path bullet below (#1778 찐빠 #4); (i) when the edit writes a runnable
+        command or code snippet into rule/skill/guide/workflow text, the agent MUST copy that command OUT of the
+        edited file and run it on a positive and a negative fixture, putting both raw outputs in the
+        completion report (the #1752 fixture bullet below, carried here so it reaches the prompt;
+        #1781 찐빠 #1).
       - Any implement-stage delegation that writes an executable command or code snippet into
         rule/skill/guide/workflow TEXT (not a prose-only edit) MUST require, as a completion-report
         item, that the agent copy the command OUT of the edited file and run it once on a positive
@@ -706,6 +713,13 @@ steps:
         measurement; (c) no working around a discovered product or measurement defect to force
         a pass — halt and report instead (R023 「Conditional-Output Verification」 양성/음성 짝
         원칙 확장, #1721 찐빠 #1, #1727 찐빠 #2, #1728 찐빠 #2).
+      - A delegation that adds an item to an existing list or loop (e.g. a sync-file list) MUST
+        require, as a completion item, citing with file:line how that list handles preserveFiles,
+        protection and backup contracts, and judging whether the new item is consistent with the
+        other components in the same run (#1786 찐빠 #1).
+      - A delegation that writes migration/upgrade tests MUST require, as a completion item, that
+        every "old state" fixture be derived from that version's generator output (or real tagged
+        artifacts; cite the call used), never hand-written from source expressions (#1786 찐빠 #2).
       - Delegations (research or implementation) that design or fix test isolation or module
         mocking MUST include in their completion criteria a before/after coverage comparison
         of the target src file(s) AND the All-files total, measured with the SAME harness

package/templates/.claude/skills/playwright-compress/SKILL.md

@@ -50,14 +50,14 @@ MCP tool response (37K+ chars)
 
 ## Hook Configuration
 
-Configured in `.claude/hooks/hooks.json` PostToolUse section:
+Configured in `.claude/hooks/hooks.json` PostToolUse section (the entry whose command runs `playwright-compress.sh`):
 
 ```json
 {
   "matcher": "mcp_tool_name matches \"mcp__playwright__.*\" || mcp_tool_name matches \"mcp__claude-in-chrome__.*\"",
   "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"
 }

package/templates/guides/middleware-patterns/README.md

@@ -348,13 +348,13 @@ fi
 exit 0
 ```
 
-hooks.json 등록:
+hooks.json 등록 (illustrative — `your-middleware.sh` is a placeholder, not a shipped script):
 ```json
 {
   "PreToolUse": [
     {
       "matcher": "Bash",
-      "command": ".claude/hooks/your-middleware.sh",
+      "command": "bash \"${CLAUDE_PROJECT_DIR:-.}/.claude/hooks/scripts/your-middleware.sh\"",
       "description": "lifecycle: wrap_tool_call — 위험 패턴 감지"
     }
   ]
@@ -394,7 +394,7 @@ limitations:
 훅 스크립트의 단위 테스트:
 ```bash
 # PreToolUse hook 직접 테스트
-CLAUDE_TOOL_NAME="Bash" CLAUDE_TOOL_INPUT="rm -rf /" .claude/hooks/your-middleware.sh
+CLAUDE_TOOL_NAME="Bash" CLAUDE_TOOL_INPUT="rm -rf /" .claude/hooks/scripts/your-middleware.sh
 echo "Exit: $?"  # 0 또는 2 확인
 ```
 

package/templates/manifest.json

@@ -1,5 +1,5 @@
 {
-  "version": "1.1.93",
+  "version": "1.1.95",
   "lastUpdated": "2026-09-03",
   "omcustomMinClaudeCode": "2.1.121",
   "omcustomMinClaudeCodeReason": "Sensitive-path direct Write/Edit on .claude/** under bypassPermissions (R010 deprecation, #1101)",

package/templates/workflows/auto-dev.yaml

@@ -413,7 +413,7 @@ steps:
 
   - name: deep-plan
     skill: deep-plan
-    description: "Research-validated implementation plan (research → plan → verify) — skipped if docs-only, integrated-analysis allowed if lite; independent of tier, MAY run in lightweight mode when compression-mode-eval's 'Cross-tier — Lightweight Skill-Mode Substitution' conditions are met. MULTI-PHASE: 스킬을 그대로 spawn하지 말고 R020 「위임 경계를 Phase 개수로 설계」에 따라 단일 목표 위임으로 분할해 순차 발주한다 (#1595 #4). 리서치·계측 위임의 완료 조건에는 결론 수치를 산출물 표에서 jq/awk로 재계산해 병기하도록 명시한다(#1707 #4). 테스트 격리·모듈 모킹 수정을 다루는 리서치·계획 위임에도 대상 src 파일과 All files의 수정 전후 커버리지를 동일 하네스로 비교하도록 완료 조건에 명시합니다(#1778 찐빠 #5, implement 스텝 Rules와 동일 요구)."
+    description: "Research-validated implementation plan (research → plan → verify) — skipped if docs-only, integrated-analysis allowed if lite; independent of tier, MAY run in lightweight mode when compression-mode-eval's 'Cross-tier — Lightweight Skill-Mode Substitution' conditions are met. MULTI-PHASE: 스킬을 그대로 spawn하지 말고 R020 「위임 경계를 Phase 개수로 설계」에 따라 단일 목표 위임으로 분할해 순차 발주한다 (#1595 #4). 리서치·계측 위임의 완료 조건에는 결론 수치를 산출물 표에서 jq/awk로 재계산해 병기하도록 명시한다(#1707 #4). 테스트 격리·모듈 모킹 수정을 다루는 리서치·계획 위임에도 대상 src 파일과 All files의 수정 전후 커버리지를 동일 하네스로 비교하도록 완료 조건에 명시합니다(#1778 찐빠 #5, implement 스텝 Rules와 동일 요구). 적용 경계 표에는 각 행마다 이 위치를 읽는 위임 유형/단계 열을 두고 결함이 발생한 단계와 일치하는지 표에서 대조합니다(#1781 찐빠 #2). 기본값·출력 리터럴을 바꾸는 계획은 그 리터럴을 단언하는 기존 테스트를 git grep으로 전수 열거해 웨이브 표의 파일 소유권에 포함합니다(git grep은 untracked 파일을 보지 못하므로 git status --porcelain도 확인합니다, #1786 찐빠 #5)."
     depends_on: plan
 
   - name: implement
@@ -486,6 +486,9 @@ steps:
         its lightweight tier substitute, applying to lite and standard tiers, and
         implement-stage delegation prompts MUST carry the relevant boundary row for each edit
         (회고 제안 묶음 릴리즈의 통합 분석에 적용 경계 표 필수, #1735 찐빠 #4).
+        The table MUST include a column naming which delegation type/step reads each target
+        location and be checked against the step where the defect occurred; a mismatch means
+        the clause must also be placed at the location that step reads (#1781 찐빠 #2).
       - Before dispatching any delegation prompt in an iteration that created or reinforced a rule
         clause, the orchestrator MUST compare that clause's Anti-pattern table rows 1:1 against the
         delegation prompt's sentences and rewrite any match (cause: the R016 「신설 조항의 동일
@@ -502,7 +505,7 @@ steps:
         block: (a) Korean 합쇼체 for new sentences, do not imitate adjacent 반말 — completion
         criteria MUST include THREE deterministic checks, scoped to edited text files (md/yaml)
         and restricted to added lines only (`git diff -U0 -- <edited md/yaml files> | grep
-        '^+'`): family check `grep -cE '(한다|된다|않는다|따른다|했다|있다|없다|넣는다|이다)[.。 "]'`
+        '^+'`): family check `grep -cE '(한다|된다|않는다|따른다|했다|있다|없다|넣는다|이다)[.。 "(]'`
         = 0, line-final auxiliary check `grep -E '다[.。"]?$' | grep -vcE '니다[.。"]?$'` = 0
         (the family list alone misses endings such as 따른다/했다/있다 and unpunctuated
         line-final endings; the line-final check excludes 합쇼체 `-니다` endings so it does not
@@ -521,7 +524,7 @@ steps:
         python3 `re` equivalent of all THREE checks above, reading from the SAME piped stdin as
         (a) (`git diff -U0 -- <edited md/yaml files> | grep '^+' | python3 -c "import re,sys;
         text=sys.stdin.read();
-        fam=len(re.findall(r'(한다|된다|않는다|따른다|했다|있다|없다|넣는다|이다)[.。 \"]', text));
+        fam=len(re.findall(r'(한다|된다|않는다|따른다|했다|있다|없다|넣는다|이다)[.。 \"(]', text));
         lf=len(re.findall(r'다[.。\"]?$', text, re.M)) - len(re.findall(r'니다[.。\"]?$', text, re.M));
         noun=len(re.findall(r'(함|됨|필요)[.。\"]?$', text, re.M));
         print(fam, lf, noun)"`, NOT a file-argument read — the input is the piped diff, never a
@@ -550,7 +553,11 @@ steps:
         delegation creates MUST live under a per-agent-unique path (`$TMPDIR` or the session
         scratchpad), never a fixed shared path (#1722 찐빠 #7). This governs the agent's own
         ad-hoc temp files; temp paths created at runtime by test code are governed by the
-        separate test-code temp-path bullet below (#1778 찐빠 #4).
+        separate test-code temp-path bullet below (#1778 찐빠 #4); (i) when the edit writes a runnable
+        command or code snippet into rule/skill/guide/workflow text, the agent MUST copy that command OUT of the
+        edited file and run it on a positive and a negative fixture, putting both raw outputs in the
+        completion report (the #1752 fixture bullet below, carried here so it reaches the prompt;
+        #1781 찐빠 #1).
       - Any implement-stage delegation that writes an executable command or code snippet into
         rule/skill/guide/workflow TEXT (not a prose-only edit) MUST require, as a completion-report
         item, that the agent copy the command OUT of the edited file and run it once on a positive
@@ -706,6 +713,13 @@ steps:
         measurement; (c) no working around a discovered product or measurement defect to force
         a pass — halt and report instead (R023 「Conditional-Output Verification」 양성/음성 짝
         원칙 확장, #1721 찐빠 #1, #1727 찐빠 #2, #1728 찐빠 #2).
+      - A delegation that adds an item to an existing list or loop (e.g. a sync-file list) MUST
+        require, as a completion item, citing with file:line how that list handles preserveFiles,
+        protection and backup contracts, and judging whether the new item is consistent with the
+        other components in the same run (#1786 찐빠 #1).
+      - A delegation that writes migration/upgrade tests MUST require, as a completion item, that
+        every "old state" fixture be derived from that version's generator output (or real tagged
+        artifacts; cite the call used), never hand-written from source expressions (#1786 찐빠 #2).
       - Delegations (research or implementation) that design or fix test isolation or module
         mocking MUST include in their completion criteria a before/after coverage comparison
         of the target src file(s) AND the All-files total, measured with the SAME harness