Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/jerrylalala/compound-engineering/codex-first-executornpx skills add Jerrylalala/compound-engineering --skill codex-first-executorgit clone --depth 1 https://github.com/Jerrylalala/compound-engineeringWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00059 | $0.01268 |
| Opus 5 | $0.00030 | $0.00634 |
| Sonnet 5 | $0.00012 | $0.00254 |
| Haiku 4.5 | $0.00006 | $0.00127 |
Grade A, and why
codex-first-executor scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 134 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Codex-first External Executor — 任务特征路由
Codex 洞察（P7，P5+P8 合并）：路由策略和执行器可靠性不可分。 这不是「Codex 优先」品牌路由，而是「按任务特征选择最合适的执行器」。
前置条件
必须先通过 Executor Capability Gate 检查（skills-custom/executor-capability-gate）再执行路由决策。
任务特征路由矩阵（Dispatch Policy Matrix）
| 任务特征 | 推荐执行器 | 理由 |
|---|---|---|
| 高风险改动（auth、payment、migration、production data） | Claude 主做 + ce:review | 需要项目上下文和谨慎判断 |
| 大量机械 patch（格式化、重命名、批量模板替换） | Codex | 重复性操作，Codex 擅长 |
| 前端视觉任务（UI 组件、CSS） | Claude + design contract + UI Review Contract | 需要视觉理解 |
| 纯分析/代码审查 | Codex 交叉审核（[C] 参数） | 补充视角，有价值 |
| 跨文件重构（需要项目上下文） | Claude 主做 | 上下文依赖强 |
| 独立算法实现（无上下文依赖） | Codex（可选） | 相对独立，Codex 可胜任 |
| 文档写作/注释 | Codex（可选） | 相对机械 |
执行器可靠性管理
重试策略
调用 Codex
├─ 成功（exit 0）→ 使用结果
├─ 失败（exit 1, 网络错误）→ 等待 30s，重试 1 次
├─ 重试失败 → 降级到 Claude 执行
└─ Rate limit (429) → 等待 retry-after header 时间，或降级
超时管理
# Codex 调用超时（180s）
timeout 180 codex -- "$TASK_PROMPT"
EXIT_CODE=$?
if [ $EXIT_CODE -eq 124 ]; then
echo "Codex 超时，降级到 Claude 执行"
elif [ $EXIT_CODE -eq 0 ]; then
# 成功调用后记录时间（供 executor-capability-gate Check 4 Rate Limit 使用）
date +%s > ~/.codex/.last_call
fi
降级路径
Codex → Claude (降级)
触发条件：CLI 不可用 / 超时 / rate limit / 任务不适合
降级后：记录降级原因到 state.md（如有 Task Bundle）
Codex 审核模式（[C] 参数）
当用户使用 ce:review [C] 或 ce:brainstorm [C] 时，Codex 作为交叉审核视角，不是主执行器：
Claude 执行主任务
↓
Claude 生成结果/审核报告
↓
Codex 独立审核（仅审核，不修改代码）
↓
多模型仲裁（Review Contract 定义的权重）
├─ Claude trust_score: 1.0
└─ Codex trust_score: 0.85
↓
合并结论（多方一致 → confidence: high）
Gemini 策略
架构澄清：Gemini 有两种使用模式，定位不同：
| 模式 | 状态 | 入口 |
|---|---|---|
| 审核视角（交叉审核，不执行代码） | ✅ 已激活 | ce:review [G]、ce:brainstorm [G] |
| 执行器路由（Gemini 作为主执行引擎） | ⏸ 暂缓 | 无 |
当前决策：
[G]审核参数已可用：Gemini 作为独立视角交叉审核 PR 或方案，结果整合进 review/brainstorm 报告- Gemini 作为主线执行器（类似 Codex 替代 Claude 执行任务）暂不实现
- 本 Executor 路由矩阵目前只路由 Claude ↔ Codex，不包含 Gemini 执行路径
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 134 lines · 59 tokens per session scan A 49ce6a58f0dd
codex-first-executor is a skill published in the GitHub repository Jerrylalala/compound-engineering (5 stars, last pushed 3mo ago), licensed MIT. It adds 59 tokens to every session and 1,268 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
auto-perf-optimize
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.
chat-perf
Run chat perf benchmarks and memory leak checks against the local dev build or any published VS Code version. Use when investigating chat rendering regressions, validating perf-sensitive changes to chat UI, or checking for memory leaks in the chat response pipeline.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…