Codex System — Deep Reasoning Partner

Codex CLI (gpt-5.2-codex) is your highly capable supporter for deep reasoning tasks.

상세규칙: .claude/rules/codex-delegation.md

Context Management (CRITICAL)

서브에이전트 경유 권장이다. 메인 오케스트레이터의 컨텍스트를 절약하기 위해.

| 상황 | 방법 | |------|------| | 자세한 설계 상담 | 서브 에이전트를 통해 (권장) | | 디버그 분석 | 서브 에이전트를 통해 (권장) | | 짧은 질문 (1-2 문 답변) | 직접 호출 확인 |

When to Consult (MUST)

| Situation | Trigger Examples | |-----------|------------------| | Design decisions | "어디 디자인?" "아키텍처" / "How to design?" | | Debugging | "왜 움직이지 않는가?" "오류" / "Debug" "Error" | | Trade-off analysis | "어느 쪽이 좋은가?" "비교해" / "Compare" "Which?" | | Complex implementation | "구현 방법" "어떻게 만드는가?" / "How to implement?" | | Refactoring | "리팩터" "간단하게" / "Refactor" "Simplify" | | Code review | "리뷰하고" "확인하고" / "Review" "Check" |

When NOT to Consult

• Simple file edits, typo fixes
• Following explicit user instructions
• git commit, running tests, linting
• Tasks with obvious single solutions

How to Consult

Recommended: Subagent Pattern

Use Task tool with subagent_type='general-purpose' to preserve main context.

Task tool parameters:
- subagent_type: "general-purpose"
- run_in_background: true (optional, for parallel work)
- prompt: |
    Consult Codex about: {topic}

    codex exec --model gpt-5.2-codex --sandbox read-only --full-auto "
    {question for Codex}
    " 2>/dev/null

    Return CONCISE summary (key recommendation + rationale).

Direct Call (Short Questions Only)

For quick questions expecting 1-2 sentence answers:

codex exec --model gpt-5.2-codex --sandbox read-only --full-auto "Brief question" 2>/dev/null

Workflow (Subagent)

1. Spawn subagent with Codex consultation prompt
2. Continue your work → Subagent runs in parallel
3. Receive summary → Subagent returns concise insights

Sandbox Modes

| Mode | Use Case | |------|----------| | read-only | Analysis, review, debugging advice | | workspace-write | Implementation, refactoring, fixes |

Language Protocol

1. Ask Codex in English
2. Receive response in English
3. Execute based on advice (or let Codex execute)
4. Report to user in Korean

Task Templates

Design Review

codex exec --model gpt-5.2-codex --sandbox read-only --full-auto "
Review this design approach for: {feature}

Context:
{relevant code or architecture}

Evaluate:
1. Is this approach sound?
2. Alternative approaches?
3. Potential issues?
4. Recommendations?
" 2>/dev/null

Debug Analysis

codex exec --model gpt-5.2-codex --sandbox read-only --full-auto "
Debug this issue:

Error: {error message}
Code: {relevant code}
Context: {what was happening}

Analyze root cause and suggest fixes.
" 2>/dev/null

Code Review

See: references/code-review-task.md

Refactoring

See: references/refactoring-task.md

Integration with Gemini

| Task | Use | |------|-----| | Need research first | Gemini → then Codex | | Design decision | Codex directly | | Library comparison | Gemini research → Codex decision |

Why Codex?

• Deep reasoning: Complex analysis and problem-solving
• Code expertise: Implementation strategies and patterns
• Consistency: Same project context via context-loader skill
• Parallel work: Background execution keeps you productive