triflux 10.0.0 → 10.0.2

package/skills/tfx-autoresearch/SKILL.md

@@ -2,16 +2,17 @@
 name: tfx-autoresearch
 description: >
   자율 웹 리서치 → 실행 가능 계획 도출.
-  주제를 받아 웹 검색 → 정보 수집 → 분석 → 구조화된 리서치 보고서 + 실행 계획 생성.
-  '자율 리서치', 'autoresearch', '검색하고 정리해', 'research and plan' 같은 요청에 사용.
-  단순 검색은 tfx-research, 자율 리서치+계획 도출은 이 스킬.
+  주제를 받아 웹 검색 → 정보 수집 → 분석 → 구조화된 리서치 보고서 생성.
+  '알아봐', '조사해', '리서치해줘', '검색하고 정리해', 'research and plan' 같은 요청에 반드시 사용.
+  웹 검색 후 실행 가능한 계획까지 도출이 필요할 때 적극 활용.
 triggers:
   - autoresearch
   - 자율 리서치
   - 자동 리서치
   - 웹 리서치
-  - research and plan
-  - 검색하고 정리해
+  - 조사해
+  - 알아봐
+  - research this
 argument-hint: "<topic>"
 source: GAP 분석 P2 #8
 ---

package/skills/tfx-codex/SKILL.md.tmpl

@@ -8,10 +8,9 @@ triggers:
 argument-hint: "\"작업 설명\" | N:codex \"작업 설명\""
 ---
 
-# tfx-codex — Codex-Only 오케스트레이터
+# {{SKILL_NAME}} — Codex-Only 오케스트레이터
 
-> **ARGUMENTS 처리**: 이 스킬이 `ARGUMENTS: <값>`과 함께 호출되면, 해당 값을 사용자 입력으로 취급하여
-> 워크플로우의 첫 단계 입력으로 사용한다. ARGUMENTS가 비어있거나 없으면 기존 절차대로 사용자에게 입력을 요청한다.
+{{> base}}
 
 
 > **래퍼**: tfx-auto의 Codex 전용 바로가기. TFX_CLI_MODE=codex.

package/skills/tfx-codex-swarm-workspace/iteration-1/benchmark.json

@@ -0,0 +1,33 @@
+{
+  "skill_name": "tfx-codex-swarm",
+  "iteration": 1,
+  "configurations": [
+    {
+      "name": "with_skill",
+      "pass_rate": {"mean": 1.0, "stddev": 0.0},
+      "total_tokens": {"mean": 30953, "stddev": 6089},
+      "duration_seconds": {"mean": 166.8, "stddev": 10.2},
+      "per_eval": [
+        {"eval_name": "full-swarm-all-prds", "pass_rate": 1.0, "assertions_passed": 7, "assertions_total": 7, "tokens": 37789, "duration_s": 178.4},
+        {"eval_name": "selective-spawn-with-override", "pass_rate": 1.0, "assertions_passed": 4, "assertions_total": 4, "tokens": 29085, "duration_s": 163.1},
+        {"eval_name": "implicit-swarm-no-keywords", "pass_rate": 1.0, "assertions_passed": 5, "assertions_total": 5, "tokens": 25984, "duration_s": 158.8}
+      ]
+    },
+    {
+      "name": "without_skill",
+      "pass_rate": {"mean": 0.875, "stddev": 0.144},
+      "total_tokens": {"mean": 53786, "stddev": 10873},
+      "duration_seconds": {"mean": 264.4, "stddev": 49.3},
+      "per_eval": [
+        {"eval_name": "full-swarm-all-prds", "pass_rate": 1.0, "assertions_passed": 7, "assertions_total": 7, "tokens": 50555, "duration_s": 218.2},
+        {"eval_name": "selective-spawn-with-override", "pass_rate": 1.0, "assertions_passed": 4, "assertions_total": 4, "tokens": 65928, "duration_s": 259.5},
+        {"eval_name": "implicit-swarm-no-keywords", "pass_rate": 0.6, "assertions_passed": 3, "assertions_total": 5, "tokens": 44875, "duration_s": 315.5}
+      ]
+    }
+  ],
+  "delta": {
+    "pass_rate": "+0.125 (87.5% → 100%)",
+    "tokens": "-42.5% (53.8K → 31.0K avg)",
+    "duration": "-36.9% (264s → 167s avg)"
+  }
+}

package/skills/tfx-codex-swarm-workspace/iteration-1/full-swarm-all-prds/eval_metadata.json

@@ -0,0 +1,42 @@
+{
+  "eval_id": 1,
+  "eval_name": "full-swarm-all-prds",
+  "prompt": ".omx/plans/ 에 PRD 7개 있는데 전부 codex swarm으로 돌려줘. 구현은 autopilot, 조사는 plan만 쓰고. 하나의 WT에 탭으로 넣어줘",
+  "assertions": [
+    {
+      "name": "all-7-prds-found",
+      "description": "7개 PRD 파일 모두 스캔 및 나열",
+      "check": "classification.md에 issue 24~30 모든 PRD가 언급되어야 함"
+    },
+    {
+      "name": "correct-type-classification",
+      "description": "구현(24,26,27,28,29) vs 조사(25,30) 정확히 분류",
+      "check": "issue 25, 30은 조사/investigation으로, 나머지는 구현/implement로 분류"
+    },
+    {
+      "name": "profile-routing-varies",
+      "description": "조사 태스크에 gpt54, 구현에 codex53 프로파일 적용",
+      "check": "routing.md에서 조사=gpt54_*, 구현=codex53_* 매핑"
+    },
+    {
+      "name": "worktree-per-task",
+      "description": "7개 독립 git worktree 생성 명령 존재",
+      "check": "commands.sh에 git worktree add 명령이 7개"
+    },
+    {
+      "name": "psmux-sessions",
+      "description": "7개 psmux 세션 생성 + codex 실행 명령",
+      "check": "commands.sh에 psmux new-session이 7개, send-keys에 codex 명령 포함"
+    },
+    {
+      "name": "single-wt-window",
+      "description": "하나의 WT 윈도우에 7탭 attach",
+      "check": "commands.sh에 wt-new-window 1회 + wt-tab 6회 (또는 wt.exe -w new + new-tab)"
+    },
+    {
+      "name": "omx-skill-mapping",
+      "description": "구현=$autopilot, 조사=$plan 매핑",
+      "check": "classification.md 또는 commands.sh에서 구현 태스크에 $autopilot, 조사에 $plan"
+    }
+  ]
+}

package/skills/tfx-codex-swarm-workspace/iteration-1/full-swarm-all-prds/with_skill/grading.json

@@ -0,0 +1,11 @@
+{
+  "expectations": [
+    {"text": "all-7-prds-found", "passed": true, "evidence": "classification.md에 issue 24~30 모두 나열됨"},
+    {"text": "correct-type-classification", "passed": true, "evidence": "25,30=investigate, 24,26,27,28,29=implement 정확"},
+    {"text": "profile-routing-varies", "passed": true, "evidence": "implement→codex53_high, investigate→gpt54_high"},
+    {"text": "worktree-per-task", "passed": true, "evidence": "commands.sh에 git worktree add 7개 확인"},
+    {"text": "psmux-sessions", "passed": true, "evidence": "7개 psmux new-session + send-keys with codex"},
+    {"text": "single-wt-window", "passed": true, "evidence": "wt-new-window 1회 + wt-tab 6회"},
+    {"text": "omx-skill-mapping", "passed": true, "evidence": "implement=$plan→$autopilot, investigate=$plan"}
+  ]
+}

package/skills/tfx-codex-swarm-workspace/iteration-1/full-swarm-all-prds/with_skill/outputs/analysis.md

@@ -0,0 +1,87 @@
+# Codex Swarm - Summary Analysis
+
+> Generated: 2026-03-30
+> Skill: `tfx-codex-swarm` (SKILL.md)
+> Input: `.omx/plans/prd-issue-{24..30}.md` (7 PRDs)
+> User request: "전부 codex swarm으로 돌려줘. 구현은 autopilot, 조사는 plan만 쓰고. 하나의 WT에 탭으로 넣어줘"
+
+## Scan Result (Step 1)
+
+`.omx/plans/` contains 20 files total. Among them, 7 are PRD files matching `prd-issue-*.md`:
+
+| # | File | Lines |
+|---|------|-------|
+| 24 | `prd-issue-24-remote-spawn-file-transfer.md` | 73 |
+| 25 | `prd-issue-25-remote-spawn-resize-blank-screen.md` | 76 |
+| 26 | `prd-issue-26-remote-spawn-session-cleanup.md` | 70 |
+| 27 | `prd-issue-27-hud-dashboard-anchor.md` | 73 |
+| 28 | `prd-issue-28-headless-guard-spawn-deadlock.md` | 69 |
+| 29 | `prd-issue-29-headless-cwd-propagation.md` | 70 |
+| 30 | `prd-issue-30-remote-spawn-terminal-minimize.md` | 70 |
+
+The remaining 13 files are plans (`plan-issue-*`), autopilot specs/impls, and an issue-validity report -- not PRDs.
+
+## Selection (Step 2)
+
+User said "전부" (all) -- skipped interactive selection. All 7 PRDs selected.
+
+## Classification (Step 3)
+
+| Type | Count | Issues | OMX Skill |
+|------|-------|--------|-----------|
+| **implement** | 5 | 24, 26, 27, 28, 29 | `$plan` -> `$autopilot` |
+| **investigate** | 2 | 25, 30 | `$plan` only |
+
+User explicitly overrode default skill assignment:
+- Implement tasks use `$autopilot` (not the default `$plan -> $autopilot` vs `$plan -> $ralph` choice)
+- Investigate tasks use `$plan` only (no execution phase)
+
+## Routing (Step 4)
+
+All 7 PRDs are size **L** (40-80 lines, 3-5 affected files, 0 high-cost keywords).
+
+| Profile | Model | Effort | Count | Issues |
+|---------|-------|--------|-------|--------|
+| `codex53_high` | gpt-5.3-codex | high | 5 | 24, 26, 27, 28, 29 |
+| `gpt54_high` | gpt-5.4 | high | 2 | 25, 30 |
+
+Investigation tasks use `gpt-5.4` (better for analysis/reasoning) instead of `gpt-5.3-codex` (optimized for code generation).
+
+## Infrastructure (Steps 5-8)
+
+### Worktrees (Step 5)
+- 7 independent git worktrees under `.codex-swarm/wt-issue-{24..30}`
+- Each branched as `codex/issue-{N}` from current HEAD
+
+### Prompts (Step 6)
+- 7 prompt files at `.codex-swarm/prompts/prompt-{24..30}.md`
+- Implement prompts: read PRD -> `$plan` -> `$autopilot`
+- Investigate prompts: read PRD -> `$plan` -> stop (no implementation)
+
+### Sessions (Step 7)
+- 7 psmux sessions: `codex-swarm-{24..30}`
+- Each runs `codex` with profile-specific flags + `--full-auto`
+
+### WT Layout (Step 8)
+- Single WT window with 7 tabs
+- First session opens new window, remaining 6 attach as tabs
+- Fallback `wt.exe` command included for non-psmux-attach environments
+
+## Final Status Table (Step 9)
+
+| # | Task | Type | OMX Skill | Profile | Worktree | Session |
+|---|------|------|-----------|---------|----------|---------|
+| 24 | file-transfer | implement | `$plan` -> `$autopilot` | `codex53_high` | `wt-issue-24` | `codex-swarm-24` |
+| 25 | resize-blank | investigate | `$plan` | `gpt54_high` | `wt-issue-25` | `codex-swarm-25` |
+| 26 | session-cleanup | implement | `$plan` -> `$autopilot` | `codex53_high` | `wt-issue-26` | `codex-swarm-26` |
+| 27 | hud-dashboard-anchor | implement | `$plan` -> `$autopilot` | `codex53_high` | `wt-issue-27` | `codex-swarm-27` |
+| 28 | guard-deadlock | implement | `$plan` -> `$autopilot` | `codex53_high` | `wt-issue-28` | `codex-swarm-28` |
+| 29 | cwd-propagation | implement | `$plan` -> `$autopilot` | `codex53_high` | `wt-issue-29` | `codex-swarm-29` |
+| 30 | terminal-minimize | investigate | `$plan` | `gpt54_high` | `wt-issue-30` | `codex-swarm-30` |
+
+## Notes
+
+- **Dry run only** -- no commands were executed. All generated commands are in `commands.sh`.
+- The investigate tasks (25, 30) will produce plans but not code changes, as the user requested.
+- All 7 sessions share the same WT window via tabs, per user request.
+- Cleanup after completion: `psmux kill-session --name "codex-swarm-*"` + `git worktree prune` + `rm -rf .codex-swarm/`

package/skills/tfx-codex-swarm-workspace/iteration-1/full-swarm-all-prds/with_skill/outputs/classification.md

@@ -0,0 +1,35 @@
+# Codex Swarm - PRD Classification Table
+
+> Generated: 2026-03-30
+> Source: `.omx/plans/prd-issue-{24..30}.md` (7 PRDs)
+> User directive: "구현은 autopilot, 조사는 plan만 쓰고"
+
+## Classification
+
+| # | PRD File | Type | Keywords Found | User Override | OMX Skill |
+|---|----------|------|----------------|---------------|-----------|
+| 24 | `prd-issue-24-remote-spawn-file-transfer.md` | **implement** (구현) | "추가", "변경" | autopilot | `$plan` -> `$autopilot` |
+| 25 | `prd-issue-25-remote-spawn-resize-blank-screen.md` | **investigate** (조사) | "조사", "investigation", "재현", "reproduce" | plan only | `$plan` |
+| 26 | `prd-issue-26-remote-spawn-session-cleanup.md` | **implement** (구현) | "추가", "변경" | autopilot | `$plan` -> `$autopilot` |
+| 27 | `prd-issue-27-hud-dashboard-anchor.md` | **implement** (구현) | "변경", "개선" | autopilot | `$plan` -> `$autopilot` |
+| 28 | `prd-issue-28-headless-guard-spawn-deadlock.md` | **implement** (구현) | "변경", "fix" | autopilot | `$plan` -> `$autopilot` |
+| 29 | `prd-issue-29-headless-cwd-propagation.md` | **implement** (구현) | "변경" | autopilot | `$plan` -> `$autopilot` |
+| 30 | `prd-issue-30-remote-spawn-terminal-minimize.md` | **investigate** (조사) | "조사", "investigation", "재현", "reproduce" | plan only | `$plan` |
+
+## Classification Logic
+
+Per SKILL.md Step 3:
+
+- **implement**: PRD content contains "구현", "implement", "추가", "변경", "fix"
+- **investigate**: PRD content contains "조사", "investigation", "재현", "reproduce"
+- **refactor**: PRD content contains "리팩터", "refactor", "정리", "개선"
+
+User explicitly overrode skill assignment:
+- All **implement** tasks -> `$plan` -> `$autopilot` (not `$ralph`)
+- All **investigate** tasks -> `$plan` only (no follow-up execution)
+
+## Summary
+
+- **5 implement** tasks: issues 24, 26, 27, 28, 29
+- **2 investigate** tasks: issues 25, 30
+- **0 refactor** tasks

package/skills/tfx-codex-swarm-workspace/iteration-1/full-swarm-all-prds/with_skill/outputs/commands.sh

@@ -0,0 +1,275 @@
+#!/usr/bin/env bash
+# =============================================================================
+# Codex Swarm — Full Swarm: 7 PRDs (Issues 24-30)
+# Generated: 2026-03-30
+# User directive: 구현=autopilot, 조사=plan만, 하나의 WT에 탭으로
+# =============================================================================
+# DO NOT EXECUTE — this is a dry-run generation for review.
+# =============================================================================
+
+set -euo pipefail
+
+PROJECT_ROOT="C:/Users/SSAFY/Desktop/Projects/cli/triflux"
+SWARM_DIR="$PROJECT_ROOT/.codex-swarm"
+WT_DIR="$SWARM_DIR/wt-issue"
+PROMPT_DIR="$SWARM_DIR/prompts"
+BASH_EXE='C:\Program Files\Git\bin\bash.exe'
+
+# =============================================================================
+# Step 5: Create git worktrees
+# =============================================================================
+
+echo "=== Step 5: Creating git worktrees ==="
+
+git worktree add "$SWARM_DIR/wt-issue-24" -b codex/issue-24 2>/dev/null || \
+git worktree add "$SWARM_DIR/wt-issue-24" codex/issue-24
+
+git worktree add "$SWARM_DIR/wt-issue-25" -b codex/issue-25 2>/dev/null || \
+git worktree add "$SWARM_DIR/wt-issue-25" codex/issue-25
+
+git worktree add "$SWARM_DIR/wt-issue-26" -b codex/issue-26 2>/dev/null || \
+git worktree add "$SWARM_DIR/wt-issue-26" codex/issue-26
+
+git worktree add "$SWARM_DIR/wt-issue-27" -b codex/issue-27 2>/dev/null || \
+git worktree add "$SWARM_DIR/wt-issue-27" codex/issue-27
+
+git worktree add "$SWARM_DIR/wt-issue-28" -b codex/issue-28 2>/dev/null || \
+git worktree add "$SWARM_DIR/wt-issue-28" codex/issue-28
+
+git worktree add "$SWARM_DIR/wt-issue-29" -b codex/issue-29 2>/dev/null || \
+git worktree add "$SWARM_DIR/wt-issue-29" codex/issue-29
+
+git worktree add "$SWARM_DIR/wt-issue-30" -b codex/issue-30 2>/dev/null || \
+git worktree add "$SWARM_DIR/wt-issue-30" codex/issue-30
+
+# =============================================================================
+# Step 6: Generate prompts + copy PRD files into worktrees
+# =============================================================================
+
+echo "=== Step 6: Generating prompts ==="
+
+mkdir -p "$PROMPT_DIR"
+
+# --- Issue 24: implement -> $plan -> $autopilot ---
+cp "$PROJECT_ROOT/.omx/plans/prd-issue-24-remote-spawn-file-transfer.md" \
+   "$SWARM_DIR/wt-issue-24/.omx/plans/prd-issue-24-remote-spawn-file-transfer.md" 2>/dev/null || \
+   { mkdir -p "$SWARM_DIR/wt-issue-24/.omx/plans" && \
+     cp "$PROJECT_ROOT/.omx/plans/prd-issue-24-remote-spawn-file-transfer.md" \
+        "$SWARM_DIR/wt-issue-24/.omx/plans/prd-issue-24-remote-spawn-file-transfer.md"; }
+
+cat > "$PROMPT_DIR/prompt-24.md" << 'PROMPT_EOF'
+triflux 프로젝트의 태스크를 구현해야 합니다.
+
+태스크 파일을 먼저 읽으세요: .omx/plans/prd-issue-24-remote-spawn-file-transfer.md
+
+작업 순서:
+1. 태스크 파일을 읽고 요구사항을 파악하세요
+2. $plan 스킬로 구현 계획을 수립하세요
+3. 계획이 완료되면 $autopilot 스킬로 자율 구현을 시작하세요
+
+프로젝트 정보:
+- triflux: Claude Code용 멀티모델 CLI 오케스트레이터
+- 언어: JavaScript/ESM (Node.js), 테스트: npm test
+PROMPT_EOF
+
+# --- Issue 25: investigate -> $plan only ---
+cp "$PROJECT_ROOT/.omx/plans/prd-issue-25-remote-spawn-resize-blank-screen.md" \
+   "$SWARM_DIR/wt-issue-25/.omx/plans/prd-issue-25-remote-spawn-resize-blank-screen.md" 2>/dev/null || \
+   { mkdir -p "$SWARM_DIR/wt-issue-25/.omx/plans" && \
+     cp "$PROJECT_ROOT/.omx/plans/prd-issue-25-remote-spawn-resize-blank-screen.md" \
+        "$SWARM_DIR/wt-issue-25/.omx/plans/prd-issue-25-remote-spawn-resize-blank-screen.md"; }
+
+cat > "$PROMPT_DIR/prompt-25.md" << 'PROMPT_EOF'
+triflux 프로젝트의 태스크를 조사해야 합니다.
+
+태스크 파일을 먼저 읽으세요: .omx/plans/prd-issue-25-remote-spawn-resize-blank-screen.md
+
+작업 순서:
+1. 태스크 파일을 읽고 요구사항을 파악하세요
+2. $plan 스킬로 조사 계획을 수립하세요
+3. 계획 수립까지만 진행하세요 (구현하지 마세요)
+
+프로젝트 정보:
+- triflux: Claude Code용 멀티모델 CLI 오케스트레이터
+- 언어: JavaScript/ESM (Node.js), 테스트: npm test
+PROMPT_EOF
+
+# --- Issue 26: implement -> $plan -> $autopilot ---
+cp "$PROJECT_ROOT/.omx/plans/prd-issue-26-remote-spawn-session-cleanup.md" \
+   "$SWARM_DIR/wt-issue-26/.omx/plans/prd-issue-26-remote-spawn-session-cleanup.md" 2>/dev/null || \
+   { mkdir -p "$SWARM_DIR/wt-issue-26/.omx/plans" && \
+     cp "$PROJECT_ROOT/.omx/plans/prd-issue-26-remote-spawn-session-cleanup.md" \
+        "$SWARM_DIR/wt-issue-26/.omx/plans/prd-issue-26-remote-spawn-session-cleanup.md"; }
+
+cat > "$PROMPT_DIR/prompt-26.md" << 'PROMPT_EOF'
+triflux 프로젝트의 태스크를 구현해야 합니다.
+
+태스크 파일을 먼저 읽으세요: .omx/plans/prd-issue-26-remote-spawn-session-cleanup.md
+
+작업 순서:
+1. 태스크 파일을 읽고 요구사항을 파악하세요
+2. $plan 스킬로 구현 계획을 수립하세요
+3. 계획이 완료되면 $autopilot 스킬로 자율 구현을 시작하세요
+
+프로젝트 정보:
+- triflux: Claude Code용 멀티모델 CLI 오케스트레이터
+- 언어: JavaScript/ESM (Node.js), 테스트: npm test
+PROMPT_EOF
+
+# --- Issue 27: implement -> $plan -> $autopilot ---
+cp "$PROJECT_ROOT/.omx/plans/prd-issue-27-hud-dashboard-anchor.md" \
+   "$SWARM_DIR/wt-issue-27/.omx/plans/prd-issue-27-hud-dashboard-anchor.md" 2>/dev/null || \
+   { mkdir -p "$SWARM_DIR/wt-issue-27/.omx/plans" && \
+     cp "$PROJECT_ROOT/.omx/plans/prd-issue-27-hud-dashboard-anchor.md" \
+        "$SWARM_DIR/wt-issue-27/.omx/plans/prd-issue-27-hud-dashboard-anchor.md"; }
+
+cat > "$PROMPT_DIR/prompt-27.md" << 'PROMPT_EOF'
+triflux 프로젝트의 태스크를 구현해야 합니다.
+
+태스크 파일을 먼저 읽으세요: .omx/plans/prd-issue-27-hud-dashboard-anchor.md
+
+작업 순서:
+1. 태스크 파일을 읽고 요구사항을 파악하세요
+2. $plan 스킬로 구현 계획을 수립하세요
+3. 계획이 완료되면 $autopilot 스킬로 자율 구현을 시작하세요
+
+프로젝트 정보:
+- triflux: Claude Code용 멀티모델 CLI 오케스트레이터
+- 언어: JavaScript/ESM (Node.js), 테스트: npm test
+PROMPT_EOF
+
+# --- Issue 28: implement -> $plan -> $autopilot ---
+cp "$PROJECT_ROOT/.omx/plans/prd-issue-28-headless-guard-spawn-deadlock.md" \
+   "$SWARM_DIR/wt-issue-28/.omx/plans/prd-issue-28-headless-guard-spawn-deadlock.md" 2>/dev/null || \
+   { mkdir -p "$SWARM_DIR/wt-issue-28/.omx/plans" && \
+     cp "$PROJECT_ROOT/.omx/plans/prd-issue-28-headless-guard-spawn-deadlock.md" \
+        "$SWARM_DIR/wt-issue-28/.omx/plans/prd-issue-28-headless-guard-spawn-deadlock.md"; }
+
+cat > "$PROMPT_DIR/prompt-28.md" << 'PROMPT_EOF'
+triflux 프로젝트의 태스크를 구현해야 합니다.
+
+태스크 파일을 먼저 읽으세요: .omx/plans/prd-issue-28-headless-guard-spawn-deadlock.md
+
+작업 순서:
+1. 태스크 파일을 읽고 요구사항을 파악하세요
+2. $plan 스킬로 구현 계획을 수립하세요
+3. 계획이 완료되면 $autopilot 스킬로 자율 구현을 시작하세요
+
+프로젝트 정보:
+- triflux: Claude Code용 멀티모델 CLI 오케스트레이터
+- 언어: JavaScript/ESM (Node.js), 테스트: npm test
+PROMPT_EOF
+
+# --- Issue 29: implement -> $plan -> $autopilot ---
+cp "$PROJECT_ROOT/.omx/plans/prd-issue-29-headless-cwd-propagation.md" \
+   "$SWARM_DIR/wt-issue-29/.omx/plans/prd-issue-29-headless-cwd-propagation.md" 2>/dev/null || \
+   { mkdir -p "$SWARM_DIR/wt-issue-29/.omx/plans" && \
+     cp "$PROJECT_ROOT/.omx/plans/prd-issue-29-headless-cwd-propagation.md" \
+        "$SWARM_DIR/wt-issue-29/.omx/plans/prd-issue-29-headless-cwd-propagation.md"; }
+
+cat > "$PROMPT_DIR/prompt-29.md" << 'PROMPT_EOF'
+triflux 프로젝트의 태스크를 구현해야 합니다.
+
+태스크 파일을 먼저 읽으세요: .omx/plans/prd-issue-29-headless-cwd-propagation.md
+
+작업 순서:
+1. 태스크 파일을 읽고 요구사항을 파악하세요
+2. $plan 스킬로 구현 계획을 수립하세요
+3. 계획이 완료되면 $autopilot 스킬로 자율 구현을 시작하세요
+
+프로젝트 정보:
+- triflux: Claude Code용 멀티모델 CLI 오케스트레이터
+- 언어: JavaScript/ESM (Node.js), 테스트: npm test
+PROMPT_EOF
+
+# --- Issue 30: investigate -> $plan only ---
+cp "$PROJECT_ROOT/.omx/plans/prd-issue-30-remote-spawn-terminal-minimize.md" \
+   "$SWARM_DIR/wt-issue-30/.omx/plans/prd-issue-30-remote-spawn-terminal-minimize.md" 2>/dev/null || \
+   { mkdir -p "$SWARM_DIR/wt-issue-30/.omx/plans" && \
+     cp "$PROJECT_ROOT/.omx/plans/prd-issue-30-remote-spawn-terminal-minimize.md" \
+        "$SWARM_DIR/wt-issue-30/.omx/plans/prd-issue-30-remote-spawn-terminal-minimize.md"; }
+
+cat > "$PROMPT_DIR/prompt-30.md" << 'PROMPT_EOF'
+triflux 프로젝트의 태스크를 조사해야 합니다.
+
+태스크 파일을 먼저 읽으세요: .omx/plans/prd-issue-30-remote-spawn-terminal-minimize.md
+
+작업 순서:
+1. 태스크 파일을 읽고 요구사항을 파악하세요
+2. $plan 스킬로 조사 계획을 수립하세요
+3. 계획 수립까지만 진행하세요 (구현하지 마세요)
+
+프로젝트 정보:
+- triflux: Claude Code용 멀티모델 CLI 오케스트레이터
+- 언어: JavaScript/ESM (Node.js), 테스트: npm test
+PROMPT_EOF
+
+# =============================================================================
+# Step 7: Create psmux sessions + launch Codex
+# =============================================================================
+
+echo "=== Step 7: Creating psmux sessions and launching Codex ==="
+
+# --- Issue 24: implement | codex53_high ---
+psmux new-session --name "codex-swarm-24" --dir "$SWARM_DIR/wt-issue-24"
+psmux send-keys --target "codex-swarm-24:0" \
+  "codex -c 'model=\"gpt-5.3-codex\"' -c 'model_reasoning_effort=\"high\"' --full-auto \"\$(cat $PROMPT_DIR/prompt-24.md)\"" Enter
+
+# --- Issue 25: investigate | gpt54_high ---
+psmux new-session --name "codex-swarm-25" --dir "$SWARM_DIR/wt-issue-25"
+psmux send-keys --target "codex-swarm-25:0" \
+  "codex -c 'model=\"gpt-5.4\"' -c 'model_reasoning_effort=\"high\"' --full-auto \"\$(cat $PROMPT_DIR/prompt-25.md)\"" Enter
+
+# --- Issue 26: implement | codex53_high ---
+psmux new-session --name "codex-swarm-26" --dir "$SWARM_DIR/wt-issue-26"
+psmux send-keys --target "codex-swarm-26:0" \
+  "codex -c 'model=\"gpt-5.3-codex\"' -c 'model_reasoning_effort=\"high\"' --full-auto \"\$(cat $PROMPT_DIR/prompt-26.md)\"" Enter
+
+# --- Issue 27: implement | codex53_high ---
+psmux new-session --name "codex-swarm-27" --dir "$SWARM_DIR/wt-issue-27"
+psmux send-keys --target "codex-swarm-27:0" \
+  "codex -c 'model=\"gpt-5.3-codex\"' -c 'model_reasoning_effort=\"high\"' --full-auto \"\$(cat $PROMPT_DIR/prompt-27.md)\"" Enter
+
+# --- Issue 28: implement | codex53_high ---
+psmux new-session --name "codex-swarm-28" --dir "$SWARM_DIR/wt-issue-28"
+psmux send-keys --target "codex-swarm-28:0" \
+  "codex -c 'model=\"gpt-5.3-codex\"' -c 'model_reasoning_effort=\"high\"' --full-auto \"\$(cat $PROMPT_DIR/prompt-28.md)\"" Enter
+
+# --- Issue 29: implement | codex53_high ---
+psmux new-session --name "codex-swarm-29" --dir "$SWARM_DIR/wt-issue-29"
+psmux send-keys --target "codex-swarm-29:0" \
+  "codex -c 'model=\"gpt-5.3-codex\"' -c 'model_reasoning_effort=\"high\"' --full-auto \"\$(cat $PROMPT_DIR/prompt-29.md)\"" Enter
+
+# --- Issue 30: investigate | gpt54_high ---
+psmux new-session --name "codex-swarm-30" --dir "$SWARM_DIR/wt-issue-30"
+psmux send-keys --target "codex-swarm-30:0" \
+  "codex -c 'model=\"gpt-5.4\"' -c 'model_reasoning_effort=\"high\"' --full-auto \"\$(cat $PROMPT_DIR/prompt-30.md)\"" Enter
+
+# =============================================================================
+# Step 8: WT tab attach — all sessions in one window
+# =============================================================================
+
+echo "=== Step 8: Attaching all sessions to one WT window ==="
+
+# First session: new WT window
+psmux attach --session "codex-swarm-24" --wt-new-window
+
+# Remaining sessions: tabs in same window
+psmux attach --session "codex-swarm-25" --wt-tab
+psmux attach --session "codex-swarm-26" --wt-tab
+psmux attach --session "codex-swarm-27" --wt-tab
+psmux attach --session "codex-swarm-28" --wt-tab
+psmux attach --session "codex-swarm-29" --wt-tab
+psmux attach --session "codex-swarm-30" --wt-tab
+
+# --- Fallback: wt.exe direct (if psmux attach unavailable) ---
+# wt.exe -w new \
+#   --title "issue-24 file-transfer" -d "$SWARM_DIR/wt-issue-24" "$BASH_EXE" -c "psmux attach codex-swarm-24" \; \
+#   new-tab --title "issue-25 resize-blank" -d "$SWARM_DIR/wt-issue-25" "$BASH_EXE" -c "psmux attach codex-swarm-25" \; \
+#   new-tab --title "issue-26 session-cleanup" -d "$SWARM_DIR/wt-issue-26" "$BASH_EXE" -c "psmux attach codex-swarm-26" \; \
+#   new-tab --title "issue-27 hud-dashboard" -d "$SWARM_DIR/wt-issue-27" "$BASH_EXE" -c "psmux attach codex-swarm-27" \; \
+#   new-tab --title "issue-28 guard-deadlock" -d "$SWARM_DIR/wt-issue-28" "$BASH_EXE" -c "psmux attach codex-swarm-28" \; \
+#   new-tab --title "issue-29 cwd-propagation" -d "$SWARM_DIR/wt-issue-29" "$BASH_EXE" -c "psmux attach codex-swarm-29" \; \
+#   new-tab --title "issue-30 terminal-minimize" -d "$SWARM_DIR/wt-issue-30" "$BASH_EXE" -c "psmux attach codex-swarm-30"
+
+echo "=== Swarm spawned: 7 sessions in 1 WT window (7 tabs) ==="

package/skills/tfx-codex-swarm-workspace/iteration-1/full-swarm-all-prds/with_skill/outputs/routing.md

@@ -0,0 +1,56 @@
+# Codex Swarm - Profile Routing Table
+
+> Generated: 2026-03-30
+
+## Size Assessment (Step 4-1)
+
+| # | PRD File | Lines | Affected Files | High-cost Keywords | Dependencies | Size |
+|---|----------|-------|----------------|-------------------|--------------|------|
+| 24 | prd-issue-24 | 73 | 4 (`remote-spawn.mjs`, handoff, scp, staging) | 0 | none | **L** |
+| 25 | prd-issue-25 | 76 | 4 (`psmux`, SSH PTY, Claude TUI, WT) | 0 | none | **L** |
+| 26 | prd-issue-26 | 70 | 3 (`remote-spawn`, `psmux`, hub stale cleanup) | 0 | none | **L** |
+| 27 | prd-issue-27 | 73 | 4 (`wt.exe` attach, dashboard viewer, headless, config) | 0 | none | **L** |
+| 28 | prd-issue-28 | 69 | 4 (`headless-guard.mjs`, `hub/team/psmux.mjs`, codex, gemini) | 0 | none | **L** |
+| 29 | prd-issue-29 | 70 | 5 (`tfx-route.sh`, `parse-args.mjs`, `start-headless.mjs`, `headless.mjs`, `buildHeadlessCommand`) | 0 | none | **L** |
+| 30 | prd-issue-30 | 70 | 3 (`remote-spawn.mjs`, `headless.mjs`, WT) | 0 | none | **L** |
+
+### Size Thresholds (from SKILL.md)
+
+| Size | PRD Lines | Affected Files | High-cost Keywords |
+|------|-----------|----------------|-------------------|
+| XL   | 80+       | 6+             | 2+                |
+| L    | 40-80     | 3-5            | 0-1               |
+| M    | 20-40     | 1-2            | 0                 |
+| S    | <20       | 1              | 0                 |
+
+All 7 PRDs fall in the **L (standard)** range: 69-76 lines, 3-5 affected files, 0 high-cost keywords.
+
+## Profile Routing (Step 4-2)
+
+Per routing table: Type x Size -> Profile
+
+| Type \ Size | XL | L | M | S |
+|-------------|-----|-----|-----|-----|
+| **implement** | codex53_xhigh | **codex53_high** | codex53_med | codex53_low |
+| **investigate** | gpt54_high | **gpt54_high** | gpt54_low | mini54_med |
+
+## Final Routing
+
+| # | Task | Size | Type | Profile | Codex Flags |
+|---|------|------|------|---------|-------------|
+| 24 | file-transfer | L | implement | `codex53_high` | `-c 'model="gpt-5.3-codex"' -c 'model_reasoning_effort="high"'` |
+| 25 | resize-blank | L | investigate | `gpt54_high` | `-c 'model="gpt-5.4"' -c 'model_reasoning_effort="high"'` |
+| 26 | session-cleanup | L | implement | `codex53_high` | `-c 'model="gpt-5.3-codex"' -c 'model_reasoning_effort="high"'` |
+| 27 | hud-dashboard-anchor | L | implement | `codex53_high` | `-c 'model="gpt-5.3-codex"' -c 'model_reasoning_effort="high"'` |
+| 28 | guard-deadlock | L | implement | `codex53_high` | `-c 'model="gpt-5.3-codex"' -c 'model_reasoning_effort="high"'` |
+| 29 | cwd-propagation | L | implement | `codex53_high` | `-c 'model="gpt-5.3-codex"' -c 'model_reasoning_effort="high"'` |
+| 30 | terminal-minimize | L | investigate | `gpt54_high` | `-c 'model="gpt-5.4"' -c 'model_reasoning_effort="high"'` |
+
+## Execution Defaults
+
+| Setting | Value | Note |
+|---------|-------|------|
+| Mode | `--full-auto` | Sandbox auto-approve |
+| Worktree | enabled | Per-session independent worktree |
+| psmux | enabled | Session management |
+| WT | tabs in single window | User requested: "하나의 WT에 탭으로 넣어줘" |

package/skills/tfx-codex-swarm-workspace/iteration-1/full-swarm-all-prds/with_skill/timing.json

@@ -0,0 +1,5 @@
+{
+  "total_tokens": 37789,
+  "duration_ms": 178352,
+  "total_duration_seconds": 178.4
+}

package/skills/tfx-codex-swarm-workspace/iteration-1/full-swarm-all-prds/without_skill/grading.json

@@ -0,0 +1,11 @@
+{
+  "expectations": [
+    {"text": "all-7-prds-found", "passed": true, "evidence": "7개 PRD 모두 발견 및 나열"},
+    {"text": "correct-type-classification", "passed": true, "evidence": "5 implement + 2 investigate 정확"},
+    {"text": "profile-routing-varies", "passed": true, "evidence": "codex53_high/gpt54_high 분리 적용"},
+    {"text": "worktree-per-task", "passed": true, "evidence": "7개 worktree 생성 명령 존재"},
+    {"text": "psmux-sessions", "passed": true, "evidence": "7개 psmux 세션 생성됨"},
+    {"text": "single-wt-window", "passed": true, "evidence": "WT 탭 attach 포함"},
+    {"text": "omx-skill-mapping", "passed": true, "evidence": "autopilot/plan 매핑 존재"}
+  ]
+}