oh-my-codex-cli 0.1.0

package/.agent/skills/writer-memory/lib/synopsis-builder.ts

@@ -0,0 +1,339 @@
+/**
+ * Synopsis Builder - 정서 중심 시놉시스 생성기
+ *
+ * Korean writers think: emotion → relationship → event → plot
+ * NOT plot-first!
+ */
+
+import { loadMemory, saveMemory, now } from './memory-manager';
+import type { WriterMemory, Character, Relationship, Scene, SynopsisState } from './memory-manager';
+
+// === Synopsis Generation ===
+
+export function generateSynopsis(options?: {
+  protagonist?: string;
+  format?: 'full' | 'brief' | 'pitch';
+}): string | null {
+  const memory = loadMemory();
+  if (!memory) return null;
+
+  const format = options?.format || 'full';
+  const protagonist = options?.protagonist;
+
+  const attitude = extractProtagonistAttitude(memory, protagonist);
+  const relationships = extractCoreRelationships(memory, protagonist);
+  const theme = extractEmotionalTheme(memory);
+  const genreContrast = extractGenreVsEmotion(memory);
+  const aftertaste = extractEndingAftertaste(memory);
+
+  switch (format) {
+    case 'brief':
+      return formatBriefSynopsis(attitude, relationships, theme, memory);
+    case 'pitch':
+      return formatPitchSynopsis(attitude, relationships, theme, genreContrast, memory);
+    default:
+      return formatFullSynopsis(attitude, relationships, theme, genreContrast, aftertaste, memory);
+  }
+}
+
+// === 5 Essential Element Extractors ===
+
+function findProtagonist(memory: WriterMemory, name?: string): Character | null {
+  const chars = Object.values(memory.characters);
+  if (name) {
+    return chars.find(c => c.name === name || c.aliases?.includes(name)) || null;
+  }
+  return chars[0] || null;
+}
+
+export function extractProtagonistAttitude(memory: WriterMemory, protagonistName?: string): string {
+  const protagonist = findProtagonist(memory, protagonistName);
+
+  if (!protagonist) {
+    return '⚠️ 주인공 정보 없음. 캐릭터를 먼저 등록하세요.';
+  }
+
+  const parts: string[] = [];
+  if (protagonist.arc) parts.push(protagonist.arc);
+  if (protagonist.attitude) parts.push(protagonist.attitude);
+
+  if (parts.length === 0) {
+    return `⚠️ ${protagonist.name}의 태도 정보 미입력. arc와 attitude 필드를 채우세요.`;
+  }
+
+  return parts.join('. ');
+}
+
+export function extractCoreRelationships(memory: WriterMemory, protagonistName?: string): string {
+  const protagonist = findProtagonist(memory, protagonistName);
+
+  if (!protagonist) {
+    return '⚠️ 주인공 정보 없음.';
+  }
+
+  const rels = memory.relationships.filter(
+    r => r.from === protagonist.name || r.to === protagonist.name
+  );
+
+  if (rels.length === 0) {
+    return `⚠️ ${protagonist.name} 중심의 관계 정보 없음. 관계를 등록하세요.`;
+  }
+
+  return rels.map(r => {
+    const other = r.from === protagonist.name ? r.to : r.from;
+    return `${protagonist.name}-${other}: ${r.dynamic || r.type}`;
+  }).join('\n');
+}
+
+export function extractEmotionalTheme(memory: WriterMemory): string {
+  if (memory.themes.length === 0) {
+    return '⚠️ 테마 정보 없음. 작품의 정서적 주제를 입력하세요.';
+  }
+
+  return memory.themes.map(t => t.description || t.name).join('. ');
+}
+
+export function extractGenreVsEmotion(memory: WriterMemory): string {
+  const synopsis = memory.synopsis;
+  if (synopsis?.genreVsRealEmotion) {
+    return synopsis.genreVsRealEmotion;
+  }
+
+  const genre = memory.project.genre || '미지정';
+  return `장르: ${genre}. 실제 정서: 미정의. genreVsRealEmotion 필드를 입력하세요.`;
+}
+
+export function extractEndingAftertaste(memory: WriterMemory): string {
+  const synopsis = memory.synopsis;
+  if (synopsis?.endingAftertaste) {
+    return synopsis.endingAftertaste;
+  }
+
+  return '❌ 엔딩 정서 잔상 미입력. synopsis update endingAftertaste "..." 로 추가하세요.';
+}
+
+// === Synopsis State Management ===
+
+export function saveSynopsisState(state: SynopsisState): boolean {
+  const memory = loadMemory();
+  if (!memory) return false;
+
+  memory.synopsis = { ...state, lastGenerated: now() };
+  return saveMemory(memory);
+}
+
+export function loadSynopsisState(): SynopsisState | null {
+  const memory = loadMemory();
+  return memory?.synopsis || null;
+}
+
+export function updateSynopsisElement(element: keyof SynopsisState, value: string): boolean {
+  const memory = loadMemory();
+  if (!memory) return false;
+
+  memory.synopsis = memory.synopsis || {
+    protagonistAttitude: '',
+    coreRelationships: '',
+    emotionalTheme: '',
+    genreVsRealEmotion: '',
+    endingAftertaste: ''
+  };
+
+  (memory.synopsis as any)[element] = value;
+  memory.synopsis.lastGenerated = now();
+  return saveMemory(memory);
+}
+
+// === Format Functions ===
+
+export function formatFullSynopsis(
+  attitude: string,
+  relationships: string,
+  theme: string,
+  genreContrast: string,
+  aftertaste: string,
+  memory: WriterMemory
+): string {
+  const projectName = memory.project.name || '제목 미정';
+  const chars = Object.values(memory.characters);
+  const charList = chars.map(c => `- **${c.name}**: ${c.attitude || c.arc || '설명 없음'}`).join('\n');
+  const emotionFlow = memory.scenes
+    .filter(s => s.emotionTags?.length > 0)
+    .map(s => s.emotionTags[0])
+    .join(' → ') || '아직 정의되지 않음';
+
+  return `═══════════════════════════════
+시놉시스: ${projectName}
+═══════════════════════════════
+
+## 1. 주인공의 태도
+${attitude}
+
+## 2. 관계의 핵심 구도
+${relationships}
+
+## 3. 정서적 테마
+${theme}
+
+## 4. 장르와 실제 감정의 거리
+${genreContrast}
+
+## 5. 엔딩이 남기는 잔상
+${aftertaste}
+
+---
+**등장인물**:
+${charList || '(등장인물 없음)'}
+
+**장면 수**: ${memory.scenes.length}개
+
+**감정 흐름**: ${emotionFlow}
+`;
+}
+
+export function formatBriefSynopsis(
+  attitude: string,
+  relationships: string,
+  theme: string,
+  memory: WriterMemory
+): string {
+  const chars = Object.values(memory.characters);
+  const protagonist = chars[0];
+  const name = protagonist?.name || '주인공';
+
+  return `${name}은 ${attitude.split('.')[0]}. ${theme.split('.')[0]}을 통해 ${relationships.split('\n')[0] || '관계를 형성하며'} 변화한다.`;
+}
+
+export function formatPitchSynopsis(
+  attitude: string,
+  relationships: string,
+  theme: string,
+  genreContrast: string,
+  memory: WriterMemory
+): string {
+  const projectName = memory.project.name || '이 이야기';
+  const chars = Object.values(memory.characters);
+  const protagonist = chars[0];
+  const name = protagonist?.name || '주인공';
+
+  return `${projectName}는 ${attitude.split('.')[0]} ${name}이 ${theme.split('.')[0]}을 깨닫는 이야기. ${genreContrast.split('.')[0]}.`;
+}
+
+// === Checklist ===
+
+export interface ChecklistItem {
+  element: string;
+  elementKr: string;
+  status: 'complete' | 'partial' | 'missing';
+  source: string;
+  suggestion: string;
+}
+
+export function getSynopsisChecklist(memory: WriterMemory): ChecklistItem[] {
+  const chars = Object.values(memory.characters);
+  const protagonist = chars[0];
+
+  const checklist: ChecklistItem[] = [];
+
+  // 1. Protagonist Attitude
+  const hasArc = protagonist?.arc ? true : false;
+  const hasAttitude = protagonist?.attitude ? true : false;
+  checklist.push({
+    element: 'protagonistAttitude',
+    elementKr: '주인공 태도 요약',
+    status: hasArc && hasAttitude ? 'complete' : hasArc || hasAttitude ? 'partial' : 'missing',
+    source: protagonist ? `캐릭터 '${protagonist.name}'에서 추출` : '주인공 없음',
+    suggestion: hasArc && hasAttitude ? '' : 'char update <name> arc "..." attitude "..."'
+  });
+
+  // 2. Core Relationships
+  const relCount = protagonist ? memory.relationships.filter(
+    r => r.from === protagonist.name || r.to === protagonist.name
+  ).length : 0;
+  checklist.push({
+    element: 'coreRelationships',
+    elementKr: '관계 핵심 구도',
+    status: relCount >= 2 ? 'complete' : relCount === 1 ? 'partial' : 'missing',
+    source: `관계 ${relCount}개 등록됨`,
+    suggestion: relCount >= 2 ? '' : 'rel add <from> <to> <type>'
+  });
+
+  // 3. Emotional Theme
+  checklist.push({
+    element: 'emotionalTheme',
+    elementKr: '정서적 테마',
+    status: memory.themes.length > 0 ? 'complete' : 'missing',
+    source: `테마 ${memory.themes.length}개 등록됨`,
+    suggestion: memory.themes.length > 0 ? '' : 'theme add <name>'
+  });
+
+  // 4. Genre vs Emotion
+  const hasGenreContrast = memory.synopsis?.genreVsRealEmotion ? true : false;
+  checklist.push({
+    element: 'genreVsEmotion',
+    elementKr: '장르와 실제 감정의 거리',
+    status: hasGenreContrast ? 'complete' : 'missing',
+    source: hasGenreContrast ? '명시적으로 입력됨' : '미입력',
+    suggestion: hasGenreContrast ? '' : 'synopsis update genreVsRealEmotion "..."'
+  });
+
+  // 5. Ending Aftertaste
+  const hasAftertaste = memory.synopsis?.endingAftertaste ? true : false;
+  checklist.push({
+    element: 'endingAftertaste',
+    elementKr: '엔딩 정서 잔상',
+    status: hasAftertaste ? 'complete' : 'missing',
+    source: hasAftertaste ? '명시적으로 입력됨' : '미입력',
+    suggestion: hasAftertaste ? '' : 'synopsis update endingAftertaste "..."'
+  });
+
+  return checklist;
+}
+
+// === Export ===
+
+export function exportSynopsisAsMarkdown(): string {
+  const memory = loadMemory();
+  if (!memory) return '# Error: No memory found';
+
+  const synopsis = generateSynopsis({ format: 'full' });
+  if (!synopsis) return '# Error: Could not generate synopsis';
+
+  const meta = `---
+project: ${memory.project.name || 'Untitled'}
+genre: ${memory.project.genre || 'Unspecified'}
+generated: ${new Date().toISOString()}
+---
+
+`;
+
+  return meta + synopsis;
+}
+
+export function exportSynopsisAsJSON(): object {
+  const memory = loadMemory();
+  if (!memory) return { error: 'No memory found' };
+
+  const checklist = getSynopsisChecklist(memory);
+
+  return {
+    metadata: {
+      project: memory.project.name,
+      genre: memory.project.genre,
+      generated: new Date().toISOString()
+    },
+    elements: {
+      protagonistAttitude: extractProtagonistAttitude(memory),
+      coreRelationships: extractCoreRelationships(memory),
+      emotionalTheme: extractEmotionalTheme(memory),
+      genreVsEmotion: extractGenreVsEmotion(memory),
+      endingAftertaste: extractEndingAftertaste(memory)
+    },
+    checklist,
+    formats: {
+      full: generateSynopsis({ format: 'full' }),
+      brief: generateSynopsis({ format: 'brief' }),
+      pitch: generateSynopsis({ format: 'pitch' })
+    }
+  };
+}

package/.agent/skills/writer-memory/templates/synopsis-template.md

@@ -0,0 +1,46 @@
+# 시놉시스: {{PROJECT_NAME}}
+
+> 장르: {{GENRE}} | 최종 업데이트: {{DATE}}
+
+---
+
+## 1. 주인공의 태도 (Protagonist Attitude)
+
+{{PROTAGONIST_ATTITUDE}}
+
+## 2. 관계의 핵심 구도 (Core Relationships)
+
+{{CORE_RELATIONSHIPS}}
+
+## 3. 정서적 테마 (Emotional Theme)
+
+{{EMOTIONAL_THEME}}
+
+## 4. 장르와 실제 감정의 거리 (Genre vs Real Emotion)
+
+{{GENRE_VS_EMOTION}}
+
+## 5. 엔딩이 남기는 잔상 (Ending Aftertaste)
+
+{{ENDING_AFTERTASTE}}
+
+---
+
+## 부록
+
+### 등장인물
+
+{{CHARACTER_LIST}}
+
+### 장면 흐름
+
+{{SCENE_FLOW}}
+
+### 감정 궤도
+
+{{EMOTION_ARC}}
+
+---
+
+*이 시놉시스는 writer-memory 시스템에 의해 자동 생성되었습니다.*
+*플롯이 아닌 감정 설계도 기반의 시놉시스입니다.*

package/.governance/skill-lint.allowlist

@@ -0,0 +1,4 @@
+# Governance debt allowlist.
+# Format: <rule_id>:<repo-relative-path>
+legacy_slash_command:.agent/skills/help/SKILL.md
+legacy_slash_command:.agent/skills/mcp-setup/SKILL.md

package/.governance/skill-llm.allowlist

@@ -0,0 +1,4 @@
+# LLM governance debt allowlist
+# Format: <rule_id>:<skill_name>
+non_codex_runtime:doctor
+legacy_task_api:start-dev

package/AGENTS.md

@@ -0,0 +1,59 @@
+<!-- Generated: 2026-01-31 | Updated: 2026-01-31 -->
+
+# oh-my-codex
+
+Skill pack and workflow orchestration for **OpenAI Codex CLI**.
+
+**Version:** 0.1.0
+**Purpose:** Make Codex behave like a multi-agent conductor using structured skills
+
+## Purpose
+
+oh-my-codex enhances Codex with:
+- **Execution modes** (autopilot, ultrawork, ralph, ultrapilot, swarm, pipeline, ecomode)
+- **Quality workflows** (plan, review, ultraqa, security-review, tdd)
+- **Reusable skills** for research, diagnosis, and development flows
+
+## Key Files
+
+| File | Description |
+|------|-------------|
+| `README.md` | Entry point documentation |
+| `docs/CODEX.md` | Codex-specific install and usage guide |
+| `.agent/skills/` | All Codex skill definitions |
+| `scripts/install-codex.sh` | Global skill installer |
+
+## For AI Agents
+
+### Skill Invocation
+
+Codex auto-loads skills from:
+- `~/.codex/skills/<skill>/SKILL.md`
+- `<repo>/.codex/skills/<skill>/SKILL.md`
+
+When the user mentions a skill name or uses `$skill`, you should follow that skill.
+
+### Source-of-Truth Workflow (Mandatory)
+
+For this repository, **`oh-my-codex` is the single source of truth** for skills and docs.
+
+Required order:
+1. Edit and validate files in this repo first (for example `.agent/skills/**/SKILL.md`).
+2. Commit/push repository changes.
+3. Install/sync to runtime using installer scripts (for example `scripts/install-codex.sh` / `scripts/install-codex-force.sh`).
+
+Do **not** use `~/.codex/skills` as the primary editing location.
+Direct edits under `~/.codex/skills` are temporary at most, and must be immediately backported to repo before considering work complete.
+
+### Compatibility Notes
+
+Codex does **not** support Claude Code plugins, interception lifecycle APIs, or HUD. Any mention of:
+- Claude Code plugin commands
+- `.claude/` plugin cache paths
+- Claude Code-specific CLI commands
+
+…should be treated as **legacy** references from the original oh-my-claudecode.
+
+### Testing
+
+No mandatory tests. If you change scripts, run a quick shell check.

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 Yeachan Heo
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/README.md

@@ -0,0 +1,169 @@
+English | [简体中文](README.zh.md)
+
+# oh-my-codex
+
+[![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](https://opensource.org/licenses/MIT)
+
+**Codex skill pack with runtime setup tooling.**
+
+*Less prompt‑tuning, more shipping.*
+
+Inspired by **oh‑my‑claudecode**, rebuilt for **Codex architecture**, and informed by:
+**oh‑my‑opencode**, **everything‑claude‑code**.
+
+---
+
+## Install
+
+**From npm (global)**
+```bash
+npm install -g oh-my-codex-cli
+omcodex setup
+```
+
+**From npm (one-off)**
+```bash
+npx oh-my-codex-cli setup
+```
+
+**Recommended (runtime setup)**
+```bash
+npm run setup:omcodex
+```
+
+**Compatibility scripts**
+```bash
+./scripts/install-codex.sh --all
+```
+
+Force overwrite:
+```bash
+./scripts/install-codex-force.sh --all
+```
+
+Incremental (skip existing files):
+```bash
+./scripts/install-codex-incremental.sh --all
+```
+
+Installs:
+- **Skills** → `~/.codex/skills/`
+- **Role prompts** → `~/.codex/prompts/`
+- **Rules** → `~/.codex/rules/`
+- **MCP config + Plan mode** → `~/.codex/config.toml`
+
+**Project‑local**
+```bash
+./scripts/install-codex.sh --all --project
+```
+
+---
+
+## What You Get
+
+- **Execution modes**: `autopilot`, `ralph`, `ultrawork`, `ultraqa`, `ultrapilot`, `swarm`, `pipeline`, `ecomode`
+- **Planning & QA**: `plan`, `review`, `analyze`, `tdd`, `code-review`, `security-review`, `build-fix`, `verify`, `eval`
+- **Role prompt catalog**: `architect`, `planner`, `executor` (installed under `.codex/prompts/`)
+- **Learning & context**: `continuous-learning`, `strategic-compact`, `iterative-retrieval`, `verification-loop`
+- **Native subagent orchestration**: `spawn_agent` + `send_input` + `wait` + `close_agent` patterns across core skills
+- **Setup runtime**: `omcodex setup` scope-aware installer and `omcodex doctor` health checks
+- **Team state runtime**: `omcodex team start/status/advance/cancel` for staged execution state
+- **Auto skill routing**: `omcodex route "<task>"` recommends the best-fit skills with confidence
+- **Event-driven team auto-advance**: `omcodex team start "<task>" --auto` advances phases from notify events
+- **Local MCP runtime surfaces**: `omcodex_state`, `omcodex_memory`, `omcodex_trace` servers wired by setup
+- **Rules & guardrails**: coding, security, testing, performance, git workflow
+- **Plan mode enabled** (Codex 0.9+)
+
+---
+
+## Notify Positioning
+
+Codex does not provide Claude Code style interception lifecycle support (for example pre/post tool interception).
+
+In this project, extensions are event-driven and built on top of Codex notifications (`notify`), not execution interception.
+
+Use the notify extension workflow:
+
+```bash
+omcodex notify init
+omcodex notify status
+omcodex notify validate
+OMX_NOTIFY_PLUGINS=1 omcodex notify test
+```
+
+---
+
+## Codex vs Claude Code (High‑Level)
+
+| Capability | Claude Code (oh‑my‑claudecode) | Codex (oh‑my‑codex) |
+|---|---|---|
+| Skills‑based workflows | ✅ | ✅ (primary) |
+| Native subagent execution | ✅ | ✅ |
+| Plan Mode | ⚠️ plugin‑driven | ✅ native (0.9+ with config) |
+| MCP support | ✅ | ✅ (config.toml / CLI) |
+
+---
+
+## Typical Use Cases
+
+- **Ship a feature fast** → `autopilot: add OAuth login + tests`
+- **Relentless completion** → `ralph: refactor auth until tests pass`
+- **High‑throughput thinking** → `ultrawork: fix all lint + type errors`
+- **Quality loop** → `ultraqa: run tests and fix until green`
+- **Planning only** → `plan: design a scalable API for X`
+
+---
+
+## Rules (Templates)
+
+Rules are optional guardrails you can copy to `.codex/rules/`:
+- `agents.md`, `coding-style.md`, `git-workflow.md`, `notify.md`
+- `patterns.md`, `performance.md`, `security.md`, `testing.md`
+- `dev.md`, `research.md`, `review.md`
+
+Install rules automatically:
+```bash
+./scripts/install-codex.sh --rules
+```
+
+---
+
+## Skill Governance
+
+Run the governance gate before shipping skill changes:
+
+```bash
+npm run governance:skills
+npm run governance:skills:llm
+npm run eval:skills
+```
+
+This command enforces the skill documentation baseline by blocking:
+- Legacy slash command patterns (for example `Run: /verify`)
+- Plugin-only runtime instructions (for example `cc --plugin-dir`)
+- Legacy task API syntax (`Task(...)`-style examples)
+
+See `docs/SKILL_GOVERNANCE.md` for policy, blockers, and debt tracking.
+
+---
+
+
+## Quick Start
+
+```
+autopilot: build a REST API for managing tasks
+```
+
+---
+
+## Docs
+
+- `docs/CODEX.md`
+- `docs/ALIGNMENT.md`
+- `docs/NOTIFY.md`
+
+---
+
+## License
+
+MIT