consult-codex

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Dual-AI Consultation: Codex GPT-5.3 vs Code-Searcher

双AI咨询：Codex GPT-5.3 对比 Code-Searcher

You orchestrate consultation between OpenAI's Codex GPT-5.3 and Claude's code-searcher to provide comprehensive analysis with comparison.

你可以协调OpenAI的Codex GPT-5.3与Claude的code-searcher进行咨询，从而提供包含对比的全面分析。

When to Use This Skill

何时使用该Skill

High value queries:

Complex code analysis requiring multiple perspectives
Debugging difficult issues
Architecture/design questions
Code review requests
Finding specific implementations across a codebase

Lower value (single AI may suffice):

Simple syntax questions
Basic file lookups
Straightforward documentation queries

高价值查询场景：

需要多视角分析的复杂代码问题
疑难问题调试
架构/设计相关问题
代码审查请求
在代码库中查找特定实现

低价值查询场景（单个AI即可满足）：

简单语法问题
基础文件查找
直接的文档查询

Workflow

工作流程

When the user asks a code question:

当用户提出代码相关问题时：

1. Build Enhanced Prompt

1. 构建增强型提示词

Wrap the user's question with structured output requirements:

[USER_QUESTION]

=== Analysis Guidelines ===

**Structure your response with:**
1. **Summary:** 2-3 sentence overview
2. **Key Findings:** bullet points of discoveries
3. **Evidence:** file paths with line numbers (format: `file:line` or `file:start-end`)
4. **Confidence:** High/Medium/Low with reasoning
5. **Limitations:** what couldn't be determined

**Line Number Requirements:**
- ALWAYS include specific line numbers when referencing code
- Use format: `path/to/file.ext:42` or `path/to/file.ext:42-58`
- For multiple references: list each with its line number
- Include brief code snippets for key findings

**Examples of good citations:**
- "The authentication check at `src/auth/validate.ts:127-134`"
- "Configuration loaded from `config/settings.json:15`"
- "Error handling in `lib/errors.ts:45, 67-72, 98`"

用结构化输出要求包裹用户的问题：

[USER_QUESTION]

=== 分析指南 ===

**请按照以下结构组织响应：**
1. **摘要：** 2-3句话的概述
2. **关键发现：** 要点形式列出的发现内容
3. **证据：** 带行号的文件路径（格式：`file:line` 或 `file:start-end`）
4. **置信度：** 高/中/低，并说明理由
5. **局限性：** 无法确定的内容

**行号要求：**
- 引用代码时必须包含具体行号
- 使用格式：`path/to/file.ext:42` 或 `path/to/file.ext:42-58`
- 多个引用时：分别列出每个引用及其行号
- 为关键发现附上简短的代码片段

**优秀引用示例：**
- "`src/auth/validate.ts:127-134` 处的身份验证检查"
- "配置加载自 `config/settings.json:15`"
- "`lib/errors.ts:45, 67-72, 98` 中的错误处理"

2. Invoke Both Analyses in Parallel

2. 并行调用两个分析工具

Launch both simultaneously in a single message with multiple tool calls:

For Codex GPT-5.3: Use a temp file to avoid shell quoting issues:

Step 1: Write the enhanced prompt to a temp file using the Write tool:
```
Write to $CLAUDE_PROJECT_DIR/tmp/codex-prompt.txt with the ENHANCED_PROMPT content
```
Step 2: Execute Codex with the temp file and have at least 10 minute timeout as Codex can take a while to respond:

macOS:
bash
```
zsh -i -c 'codex -p readonly exec "$(cat $CLAUDE_PROJECT_DIR/tmp/codex-prompt.txt)" --json 2>&1'
```
Linux:
bash
```
bash -i -c 'codex -p readonly exec "$(cat $CLAUDE_PROJECT_DIR/tmp/codex-prompt.txt)" --json 2>&1'
```
This approach avoids all shell quoting issues regardless of prompt content.
For Code-Searcher: Use Task tool with
```
subagent_type: "code-searcher"
```
with the same enhanced prompt

This parallel execution significantly improves response time.

在单条消息中通过多个工具调用同时启动两者：

针对Codex GPT-5.3： 使用临时文件避免Shell引用问题：

步骤1： 使用Write工具将增强型提示词写入临时文件：
```
将ENHANCED_PROMPT内容写入$CLAUDE_PROJECT_DIR/tmp/codex-prompt.txt
```
步骤2： 执行Codex并设置至少10分钟超时（因为Codex响应可能较慢）：

macOS系统：
bash
```
zsh -i -c 'codex -p readonly exec "$(cat $CLAUDE_PROJECT_DIR/tmp/codex-prompt.txt)" --json 2>&1'
```
Linux系统：
bash
```
bash -i -c 'codex -p readonly exec "$(cat $CLAUDE_PROJECT_DIR/tmp/codex-prompt.txt)" --json 2>&1'
```
无论提示词内容如何，这种方法都能避免所有Shell引用问题。
针对Code-Searcher： 使用Task工具，设置
```
subagent_type: "code-searcher"
```
并传入相同的增强型提示词

并行执行可显著提升响应速度。

2a. Parse Codex

--json

Output Files (jq Recipes)

2a. 解析Codex的

--json

输出文件（jq工具脚本）

Codex CLI with

--json

typically emits newline-delimited JSON events (JSONL). Some environments may prefix lines with terminal escape sequences; these recipes strip everything before the first

and then

fromjson?

safely.

Set a variable first:

bash

FILE="/private/tmp/claude/.../tasks/<task_id>.output"   # or a symlinked *.output to agent-*.jsonl

List event types (top-level
.type
)

bash

jq -Rr 'sub("^[^{]*";"") | fromjson? | .type // empty' "$FILE" | sort | uniq -c | sort -nr

List item types (nested
.item.type
on
item.completed
)

bash

jq -Rr 'sub("^[^{]*";"") | fromjson? | select(.type=="item.completed") | .item.type? // empty' "$FILE" | sort | uniq -c | sort -nr

Extract only “reasoning” and “agent_message” text (human-readable)

bash

jq -Rr '
  sub("^[^{]*";"")
  | fromjson?
  | select(.type=="item.completed" and (.item.type? | IN("reasoning","agent_message")))
  | "===== \(.item.type) \(.item.id) =====\n\(.item.text // "")\n"
' "$FILE"

Extract just the final
agent_message
(useful for summaries)

bash

jq -Rr '
  sub("^[^{]*";"")
  | fromjson?
  | select(.type=="item.completed" and .item.type?=="agent_message")
  | .item.text // empty
' "$FILE" | tail -n 1

Build a clean JSON array for downstream tools

bash

jq -Rn '
  [inputs
   | sub("^[^{]*";"")
   | fromjson?
   | select(.type=="item.completed" and (.item.type? | IN("reasoning","agent_message")))
   | {type:.item.type, id:.item.id, text:(.item.text // "")}
  ]
' "$FILE"

Extract command executions (command + exit code), avoiding huge stdout/stderr

Codex JSON schemas vary slightly; this tries multiple common field names.

bash

jq -Rr '
  sub("^[^{]*";"")
  | fromjson?
  | select(.type=="item.completed" and .item.type?=="command_execution")
  | [
      (.item.id // ""),
      (.item.command // .item.cmd // .item.command_line // "<no command field>"),
      (.item.exit_code // .item.exitCode // "<no exit>")
    ]
  | @tsv
' "$FILE"

Discover actual fields present in
command_execution
for your environment

bash

jq -Rr '
  sub("^[^{]*";"")
  | fromjson?
  | select(.type=="item.completed" and .item.type?=="command_execution")
  | (.item | keys | @json)
' "$FILE" | head -n 5

带

--json

参数的Codex CLI通常会输出换行分隔的JSON事件（JSONL格式）。部分环境可能会在行首添加终端转义序列；以下脚本会先移除第一个

之前的所有内容，再安全执行

fromjson?

。

先设置变量：

bash

FILE="/private/tmp/claude/.../tasks/<task_id>.output"   # 或指向agent-*.jsonl的符号链接*.output

列出事件类型（顶层
.type
字段）

bash

jq -Rr 'sub("^[^{]*";"") | fromjson? | .type // empty' "$FILE" | sort | uniq -c | sort -nr

列出条目类型（
item.completed
中的嵌套
.item.type
字段）

bash

jq -Rr 'sub("^[^{]*";"") | fromjson? | select(.type=="item.completed") | .item.type? // empty' "$FILE" | sort | uniq -c | sort -nr

仅提取“reasoning”和“agent_message”文本（人类可读内容）

bash

jq -Rr '
  sub("^[^{]*";"")
  | fromjson?
  | select(.type=="item.completed" and (.item.type? | IN("reasoning","agent_message")))
  | "===== \(.item.type) \(.item.id) =====\n\(.item.text // "")\n"
' "$FILE"

仅提取最终的
agent_message
（适用于摘要）

bash

jq -Rr '
  sub("^[^{]*";"")
  | fromjson?
  | select(.type=="item.completed" and .item.type?=="agent_message")
  | .item.text // empty
' "$FILE" | tail -n 1

构建干净的JSON数组供下游工具使用

bash

jq -Rn '
  [inputs
   | sub("^[^{]*";"")
   | fromjson?
   | select(.type=="item.completed" and (.item.type? | IN("reasoning","agent_message")))
   | {type:.item.type, id:.item.id, text:(.item.text // "")}
  ]
' "$FILE"

提取命令执行记录（命令+退出码），避免过大的stdout/stderr内容

Codex的JSON架构略有差异；以下脚本会尝试匹配多个常见字段名。

bash

jq -Rr '
  sub("^[^{]*";"")
  | fromjson?
  | select(.type=="item.completed" and .item.type?=="command_execution")
  | [
      (.item.id // ""),
      (.item.command // .item.cmd // .item.command_line // "<no command field>"),
      (.item.exit_code // .item.exitCode // "<no exit>")
    ]
  | @tsv
' "$FILE"

发现你的环境中
command_execution
实际包含的字段

bash

jq -Rr '
  sub("^[^{]*";"")
  | fromjson?
  | select(.type=="item.completed" and .item.type?=="command_execution")
  | (.item | keys | @json)
' "$FILE" | head -n 5

3. Cleanup Temp Files

3. 清理临时文件

After processing the Codex response (success or failure), clean up the temp prompt file:

bash

rm -f $CLAUDE_PROJECT_DIR/tmp/codex-prompt.txt

This prevents stale prompts from accumulating and avoids potential confusion in future runs.

处理完Codex响应后（无论成功或失败），清理临时提示词文件：

bash

rm -f $CLAUDE_PROJECT_DIR/tmp/codex-prompt.txt

这可以避免旧提示词堆积，防止在后续运行中造成混淆。

4. Handle Errors

4. 错误处理

If one agent fails or times out, still present the successful agent's response
Note the failure in the comparison: "Agent X failed to respond: [error message]"
Provide analysis based on the available response

如果其中一个AI响应失败或超时，仍需展示成功响应的AI结果
在对比部分注明失败情况：“Agent X未能响应：[错误信息]”
基于可用的响应内容提供分析

5. Create Comparison Analysis

5. 创建对比分析报告

Use this exact format:

请严格使用以下格式：

Codex (GPT-5.3) Response

Codex (GPT-5.3) 响应

[Raw output from codex-cli agent]

[来自codex-cli agent的原始输出]

Code-Searcher (Claude) Response

Code-Searcher (Claude) 响应

[Raw output from code-searcher agent]

[来自code-searcher agent的原始输出]

Comparison Table

对比表格

Aspect	Codex (GPT-5.3)	Code-Searcher (Claude)
File paths	[Specific/Generic/None]	[Specific/Generic/None]
Line numbers	[Provided/Missing]	[Provided/Missing]
Code snippets	[Yes/No + details]	[Yes/No + details]
Unique findings	[List any]	[List any]
Accuracy	[Note discrepancies]	[Note discrepancies]
Strengths	[Summary]	[Summary]

维度	Codex (GPT-5.3)	Code-Searcher (Claude)
文件路径	[具体/通用/无]	[具体/通用/无]
行号	[已提供/缺失]	[已提供/缺失]
代码片段	[是/否 + 详情]	[是/否 + 详情]
独特发现	[列出内容]	[列出内容]
准确性	[标注差异]	[标注差异]
优势	[总结]	[总结]

Agreement Level

一致程度

High Agreement: Both AIs reached similar conclusions - Higher confidence in findings
Partial Agreement: Some overlap with unique findings - Investigate differences
Disagreement: Contradicting findings - Manual verification recommended

[State which level applies and explain]

高度一致： 两个AI得出相似结论 - 对发现内容的置信度更高
部分一致： 存在重叠内容但各有独特发现 - 需深入调查差异点
不一致： 结论相互矛盾 - 建议手动验证

[说明适用的一致程度并解释]

Key Differences

核心差异

Codex GPT-5.3: [unique findings, strengths, approach]
Code-Searcher: [unique findings, strengths, approach]

Codex GPT-5.3： [独特发现、优势、分析方式]
Code-Searcher： [独特发现、优势、分析方式]

Synthesized Summary

综合摘要

[Combine the best insights from both sources into unified analysis. Prioritize findings that are:

Corroborated by both agents
Supported by specific file:line citations
Include verifiable code snippets]

[整合两个来源的最佳洞见，形成统一分析。优先考虑以下发现：

两个AI均验证的内容
带有具体file:line引用的内容
包含可验证代码片段的内容]

Recommendation

建议

[Which source was more helpful for this specific query and why. Consider:

Accuracy of file paths and line numbers
Quality of code snippets provided
Completeness of analysis
Unique insights offered]

[哪个来源对本次查询更有帮助，原因是什么。请考虑：

文件路径和行号的准确性
提供的代码片段质量
分析的完整性
提供的独特洞见]

consult-codex

Original

Translation

Dual-AI Consultation: Codex GPT-5.3 vs Code-Searcher

双AI咨询：Codex GPT-5.3 对比 Code-Searcher

When to Use This Skill

何时使用该Skill

Workflow

工作流程

1. Build Enhanced Prompt

1. 构建增强型提示词

2. Invoke Both Analyses in Parallel

2. 并行调用两个分析工具

2a. Parse Codex
`--json`
Output Files (jq Recipes)

2a. 解析Codex的
`--json`
输出文件（jq工具脚本）

3. Cleanup Temp Files

3. 清理临时文件

4. Handle Errors

4. 错误处理

5. Create Comparison Analysis

5. 创建对比分析报告

Codex (GPT-5.3) Response

Codex (GPT-5.3) 响应

Code-Searcher (Claude) Response

Code-Searcher (Claude) 响应

Comparison Table

对比表格

Agreement Level

一致程度

Key Differences

核心差异

Synthesized Summary

综合摘要

Recommendation

建议

consult-codex

Original

Translation

Dual-AI Consultation: Codex GPT-5.3 vs Code-Searcher

双AI咨询：Codex GPT-5.3 对比 Code-Searcher

When to Use This Skill

何时使用该Skill

Workflow

工作流程

1. Build Enhanced Prompt

1. 构建增强型提示词

2. Invoke Both Analyses in Parallel

2. 并行调用两个分析工具

2a. Parse Codex --json Output Files (jq Recipes)

2a. 解析Codex的--json输出文件（jq工具脚本）

3. Cleanup Temp Files

3. 清理临时文件

4. Handle Errors

4. 错误处理

5. Create Comparison Analysis

5. 创建对比分析报告

Codex (GPT-5.3) Response

Codex (GPT-5.3) 响应

Code-Searcher (Claude) Response

Code-Searcher (Claude) 响应

Comparison Table

对比表格

Agreement Level

一致程度

Key Differences

核心差异

Synthesized Summary

综合摘要

Recommendation

建议

2a. Parse Codex
`--json`
Output Files (jq Recipes)

2a. 解析Codex的
`--json`
输出文件（jq工具脚本）