Menu
The review-first prompt is selected by default. You can switch to a direct command or download a local copy.
Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
npx skills add https://github.com/materialofair/oh-my-codex --skill multi-model-researchThe command stays on one line. Scroll horizontally to inspect it before copying.
Prefer a local copy? Download the files currently available to SkillsMP.
Showing SKILL.md
Based on SOC occupation classification
| name | multi-model-research |
| description | 多代理技术调研 - 使用 Codex 主线与只读 child agent 并行收集架构、实现、文档和风险视角。 |
| auto_invoke | true |
| tags | ["research","multi-agent","codex-native","no-timeout"] |
| version | 0.2.0 |
| source | fork |
| checksum | 73425cb43114bcd9aa8c4b8e527638ed77aa0385ca025a7a52a5cbeb7754fc83 |
| updated_at | 2026-05-29T03:50:00.000Z |
| intent | research |
| layer | research |
手动触发：
使用 multi-model-research 分析 <你的问题>
确认：
根据问题选择最小代理集：
spawn_agent(agent_type="explorer", message="<architecture and repo evidence task>")
spawn_agent(agent_type="reviewer", message="<risk and correctness task>")
spawn_agent(agent_type="docs-researcher", message="<official docs verification task>")
wait_agent
close_agent
常用分工：
| Agent | 关注点 | 输出 |
|---|---|---|
| explorer | 架构、代码路径、依赖关系 | 事实地图、方案适配性 |
| reviewer | 风险、安全、测试缺口、回归风险 | findings、阻塞项 |
| docs-researcher | 官方文档、版本行为、API 真实性 | 引用来源、版本约束 |
Codex 主线必须：
## ResearchPack
### Topic
[调研主题]
### Sources
- [本地文件 / 官方文档 / 命令]
### Findings
| Dimension | Finding | Evidence | Confidence |
### Options
| Option | Pros | Cons | Risks | Fit |
### Recommendation
[推荐方案 + 理由]
### Validation Plan
1. [step] -> verify: [check]
### Open Questions
[仍需确认的问题]
如果 child agent 不可用，主线 Codex 在同一回复中使用 [EXPLORER]、[REVIEWER]、、 四段完成同等分析。
[DOCS][SYNTHESIS]