ccg-workflow 1.5.1 → 1.6.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +168 -104
- package/dist/cli.mjs +13 -4
- package/dist/index.d.mts +0 -10
- package/dist/index.d.ts +0 -10
- package/dist/index.mjs +2 -2
- package/dist/shared/{ccg-workflow.DSwxPZjT.mjs → ccg-workflow.CjatSnB0.mjs} +833 -624
- package/package.json +1 -1
- package/templates/commands/agents/planner.md +2 -2
- package/templates/commands/agents/ui-ux-designer.md +2 -2
- package/templates/commands/analyze.md +1 -3
- package/templates/commands/backend.md +150 -22
- package/templates/commands/bugfix.md +1 -3
- package/templates/commands/code.md +1 -3
- package/templates/commands/debug.md +1 -3
- package/templates/commands/dev.md +27 -22
- package/templates/commands/feat.md +3 -3
- package/templates/commands/frontend.md +154 -23
- package/templates/commands/optimize.md +1 -3
- package/templates/commands/review.md +1 -3
- package/templates/commands/test.md +1 -3
- package/templates/commands/think.md +1 -3
- package/templates/commands/enhance.md +0 -52
- package/templates/config/shared-config.md +0 -87
|
@@ -2,8 +2,6 @@
|
|
|
2
2
|
description: 多模型性能优化(Codex 后端优化 + Gemini 前端优化)
|
|
3
3
|
---
|
|
4
4
|
|
|
5
|
-
> 调用语法见 `_config.md`
|
|
6
|
-
|
|
7
5
|
## 用法
|
|
8
6
|
`/optimize <优化目标>`
|
|
9
7
|
|
|
@@ -15,7 +13,7 @@ description: 多模型性能优化(Codex 后端优化 + Gemini 前端优化)
|
|
|
15
13
|
## 流程
|
|
16
14
|
|
|
17
15
|
### Phase 1: 性能基线分析
|
|
18
|
-
1. 调用 `
|
|
16
|
+
1. 调用 `mcp__ace-tool__search_context` 检索目标代码:
|
|
19
17
|
- `project_root_path`: 项目根目录绝对路径
|
|
20
18
|
- `query`: 需要优化的目标代码描述
|
|
21
19
|
2. 识别性能关键路径
|
|
@@ -2,8 +2,6 @@
|
|
|
2
2
|
description: 多模型代码审查(并行执行),无参数时自动审查 git diff
|
|
3
3
|
---
|
|
4
4
|
|
|
5
|
-
> 调用语法见 `_config.md`
|
|
6
|
-
|
|
7
5
|
## 用法
|
|
8
6
|
`/review [CODE_OR_DESCRIPTION]`
|
|
9
7
|
|
|
@@ -35,7 +33,7 @@ git status --short
|
|
|
35
33
|
|
|
36
34
|
**If arguments provided**, use the specified code/description.
|
|
37
35
|
|
|
38
|
-
Then call `
|
|
36
|
+
Then call `mcp__ace-tool__search_context` to get related context:
|
|
39
37
|
- `project_root_path`: Project root directory absolute path
|
|
40
38
|
- `query`: Description of code/files to review
|
|
41
39
|
|
|
@@ -2,8 +2,6 @@
|
|
|
2
2
|
description: 多模型测试生成(Codex 后端测试 + Gemini 前端测试)
|
|
3
3
|
---
|
|
4
4
|
|
|
5
|
-
> 调用语法见 `_config.md`
|
|
6
|
-
|
|
7
5
|
## 用法
|
|
8
6
|
`/test <测试目标>`
|
|
9
7
|
|
|
@@ -14,7 +12,7 @@ description: 多模型测试生成(Codex 后端测试 + Gemini 前端测试)
|
|
|
14
12
|
## 流程
|
|
15
13
|
|
|
16
14
|
### Phase 1: 测试分析
|
|
17
|
-
1. 调用 `
|
|
15
|
+
1. 调用 `mcp__ace-tool__search_context` 检索:
|
|
18
16
|
- `project_root_path`: 项目根目录绝对路径
|
|
19
17
|
- `query`: 需要测试的代码/功能描述
|
|
20
18
|
- 目标代码的完整实现
|
|
@@ -2,8 +2,6 @@
|
|
|
2
2
|
description: UltraThink 深度分析(双模型并行分析 + 综合见解)
|
|
3
3
|
---
|
|
4
4
|
|
|
5
|
-
> 调用语法见 `_config.md`
|
|
6
|
-
|
|
7
5
|
## 用法
|
|
8
6
|
`/think <分析主题>`
|
|
9
7
|
|
|
@@ -15,7 +13,7 @@ description: UltraThink 深度分析(双模型并行分析 + 综合见解)
|
|
|
15
13
|
## 流程
|
|
16
14
|
|
|
17
15
|
### Phase 1: 上下文收集
|
|
18
|
-
1. 调用 `
|
|
16
|
+
1. 调用 `mcp__ace-tool__search_context` 检索相关代码:
|
|
19
17
|
- `project_root_path`: 项目根目录绝对路径
|
|
20
18
|
- `query`: 需要深度分析的问题描述
|
|
21
19
|
2. 识别分析范围和关键组件
|
|
@@ -1,52 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
description: 使用 ace-tool MCP prompt-enhancer 优化 Prompt,展示原始与增强版本供确认
|
|
3
|
-
---
|
|
4
|
-
|
|
5
|
-
## Usage
|
|
6
|
-
`/enhance <PROMPT>`
|
|
7
|
-
|
|
8
|
-
## Context
|
|
9
|
-
- Original prompt: $ARGUMENTS
|
|
10
|
-
- This command enhances prompts before execution using ace-tool's enhance_prompt.
|
|
11
|
-
|
|
12
|
-
## Your Role
|
|
13
|
-
You are the **Prompt Enhancer** that optimizes user prompts for better AI task execution.
|
|
14
|
-
|
|
15
|
-
## Process
|
|
16
|
-
|
|
17
|
-
### Step 1: Enhance Prompt
|
|
18
|
-
1. Call `{{MCP_ENHANCE_TOOL}}` with:
|
|
19
|
-
- `prompt`: The original user prompt ($ARGUMENTS)
|
|
20
|
-
- `conversation_history`: Recent conversation history (5-10 turns)
|
|
21
|
-
- `project_root_path`: Current project root directory (optional)
|
|
22
|
-
2. Extract the enhanced version
|
|
23
|
-
|
|
24
|
-
### Step 2: User Confirmation (寸止)
|
|
25
|
-
**CRITICAL**: You MUST stop and show the enhanced prompt to the user.
|
|
26
|
-
|
|
27
|
-
Display format:
|
|
28
|
-
```
|
|
29
|
-
📝 原始 Prompt:
|
|
30
|
-
<original prompt>
|
|
31
|
-
|
|
32
|
-
✨ 增强后 Prompt:
|
|
33
|
-
<enhanced prompt>
|
|
34
|
-
|
|
35
|
-
---
|
|
36
|
-
**是否使用增强后的 prompt 继续执行?(Y/N)**
|
|
37
|
-
```
|
|
38
|
-
|
|
39
|
-
Wait for user confirmation before proceeding.
|
|
40
|
-
|
|
41
|
-
### Step 3: Execute (Only after confirmation)
|
|
42
|
-
If user confirms (Y):
|
|
43
|
-
- Execute the enhanced prompt as the actual task
|
|
44
|
-
- Follow appropriate workflow based on task type
|
|
45
|
-
|
|
46
|
-
If user declines (N):
|
|
47
|
-
- Ask user for modifications or use original prompt
|
|
48
|
-
|
|
49
|
-
## Notes
|
|
50
|
-
- Always show both original and enhanced versions
|
|
51
|
-
- Never auto-execute without user confirmation
|
|
52
|
-
- The enhanced prompt provides better context for multi-model collaboration
|
|
@@ -1,87 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
description: CCG 多模型协作系统 - 共享配置和调用规范
|
|
3
|
-
---
|
|
4
|
-
|
|
5
|
-
## 配置文件
|
|
6
|
-
路径: `~/.claude/.ccg/config.toml`
|
|
7
|
-
|
|
8
|
-
```toml
|
|
9
|
-
[routing]
|
|
10
|
-
mode = "smart" # smart | parallel | sequential
|
|
11
|
-
|
|
12
|
-
[routing.frontend]
|
|
13
|
-
models = ["gemini", "claude", "codex"] # 三模型并行
|
|
14
|
-
primary = "gemini"
|
|
15
|
-
strategy = "parallel"
|
|
16
|
-
|
|
17
|
-
[routing.backend]
|
|
18
|
-
models = ["codex", "claude", "gemini"] # 三模型并行
|
|
19
|
-
primary = "codex"
|
|
20
|
-
strategy = "parallel"
|
|
21
|
-
|
|
22
|
-
[routing.review]
|
|
23
|
-
models = ["codex", "gemini", "claude"] # 三模型交叉验证
|
|
24
|
-
strategy = "parallel"
|
|
25
|
-
|
|
26
|
-
[routing.prototype]
|
|
27
|
-
models = ["codex", "gemini", "claude"] # 三模型原型生成
|
|
28
|
-
strategy = "parallel"
|
|
29
|
-
```
|
|
30
|
-
|
|
31
|
-
默认值: frontend=`gemini`, backend=`codex`, prototype=`["codex","gemini","claude"]`
|
|
32
|
-
|
|
33
|
-
---
|
|
34
|
-
|
|
35
|
-
## 调用语法模板
|
|
36
|
-
|
|
37
|
-
**⚠️ 重要**: 使用 `Bash` 工具调用 `codeagent-wrapper`,**不要**使用 `/collaborating-with-codex` 或 `/collaborating-with-gemini` 等旧的 Skill 方式!
|
|
38
|
-
|
|
39
|
-
### 基础模式
|
|
40
|
-
```bash
|
|
41
|
-
codeagent-wrapper --backend <MODEL> - $PROJECT_DIR <<'EOF'
|
|
42
|
-
ROLE_FILE: ~/.claude/.ccg/prompts/<model>/<role>.md
|
|
43
|
-
|
|
44
|
-
<TASK>
|
|
45
|
-
{{任务描述}}
|
|
46
|
-
|
|
47
|
-
Context:
|
|
48
|
-
{{相关代码}}
|
|
49
|
-
</TASK>
|
|
50
|
-
|
|
51
|
-
OUTPUT: {{输出格式}}
|
|
52
|
-
EOF
|
|
53
|
-
```
|
|
54
|
-
|
|
55
|
-
**说明**: 使用 `ROLE_FILE:` 指定提示词文件路径,让子进程自己读取,避免消耗主会话 token。
|
|
56
|
-
|
|
57
|
-
### 角色映射
|
|
58
|
-
| 任务类型 | Codex 角色 | Gemini 角色 | Claude 角色 |
|
|
59
|
-
|---------|-----------|-------------|-------------|
|
|
60
|
-
| 架构/后端 | `codex/architect.md` | `gemini/analyzer.md` | `claude/architect.md` |
|
|
61
|
-
| 前端/UI | `codex/architect.md` | `gemini/frontend.md` | `claude/architect.md` |
|
|
62
|
-
| 分析 | `codex/analyzer.md` | `gemini/analyzer.md` | `claude/analyzer.md` |
|
|
63
|
-
| 审查 | `codex/reviewer.md` | `gemini/reviewer.md` | `claude/reviewer.md` |
|
|
64
|
-
| 调试 | `codex/debugger.md` | `gemini/debugger.md` | `claude/debugger.md` |
|
|
65
|
-
| 测试 | `codex/tester.md` | `gemini/tester.md` | `claude/tester.md` |
|
|
66
|
-
| 优化 | `codex/optimizer.md` | `gemini/optimizer.md` | `claude/optimizer.md` |
|
|
67
|
-
|
|
68
|
-
### 三模型差异化定位
|
|
69
|
-
| 模型 | 专长领域 | 独特价值 |
|
|
70
|
-
|------|---------|----------|
|
|
71
|
-
| **Codex** | 后端逻辑、算法、调试 | 深度后端专业知识 |
|
|
72
|
-
| **Gemini** | 前端 UI、CSS、组件设计 | 视觉设计和用户体验 |
|
|
73
|
-
| **Claude** | 全栈整合、契约设计、跨层问题 | 桥接前后端视角 |
|
|
74
|
-
|
|
75
|
-
### 输出格式
|
|
76
|
-
| 任务类型 | OUTPUT 值 |
|
|
77
|
-
|---------|----------|
|
|
78
|
-
| 原型生成 | `Unified Diff Patch ONLY. Strictly prohibit any actual modifications.` |
|
|
79
|
-
| 代码审查 | `Review comments only. No code modifications.` |
|
|
80
|
-
| 分析诊断 | `Structured analysis/diagnostic report.` |
|
|
81
|
-
|
|
82
|
-
### 执行策略
|
|
83
|
-
| 策略 | 说明 |
|
|
84
|
-
|------|------|
|
|
85
|
-
| `parallel` | `run_in_background: true` 并行调用,`TaskOutput` 收集结果 |
|
|
86
|
-
| `fallback` | 主模型失败则调用次模型 |
|
|
87
|
-
| `round-robin` | 轮询调用 |
|