mirror of
https://github.com/catlog22/Claude-Code-Workflow.git
synced 2026-02-11 02:33:51 +08:00
a944e31962
- Introduced new analysis templates for architecture, implementation patterns, performance, quality, and security. - Created detailed development templates for component creation, debugging, feature implementation, refactoring, testing, and migration planning. - Established structured documentation guidelines for root, domain, module, and sub-module levels to enhance clarity and organization. - Implemented a hierarchy analysis template to optimize project structure and documentation depth. - Updated codex-unified documentation to reflect new command structures, template usage, and best practices for autonomous development workflows.
390 lines
12 KiB
Markdown
390 lines
12 KiB
Markdown
# Parallel Execution Command Templates
|
|
|
|
**并行执行模式的完整命令示例**
|
|
|
|
## 基础并行执行模式
|
|
|
|
### 标准并行结构
|
|
```bash
|
|
# 基本并行执行模板
|
|
(
|
|
command1 &
|
|
command2 &
|
|
command3 &
|
|
wait # 等待所有并行进程完成
|
|
)
|
|
```
|
|
|
|
### 资源限制并行执行
|
|
```bash
|
|
# 限制并行进程数量
|
|
MAX_PARALLEL=3
|
|
parallel_count=0
|
|
|
|
for cmd in "${commands[@]}"; do
|
|
eval "$cmd" &
|
|
((parallel_count++))
|
|
|
|
# 达到并发限制时等待
|
|
if ((parallel_count >= MAX_PARALLEL)); then
|
|
wait
|
|
parallel_count=0
|
|
fi
|
|
done
|
|
|
|
wait # 等待剩余进程
|
|
```
|
|
|
|
## 按架构层级并行
|
|
|
|
### 前后端分离并行
|
|
```bash
|
|
# 前后端架构并行分析
|
|
(
|
|
cd src/frontend && gemini --all-files -p "@{CLAUDE.md} Frontend architecture and patterns analysis" &
|
|
cd src/backend && gemini --all-files -p "@{CLAUDE.md} Backend services and API patterns analysis" &
|
|
cd src/shared && gemini --all-files -p "@{CLAUDE.md} Shared utilities and common patterns analysis" &
|
|
wait
|
|
)
|
|
```
|
|
|
|
### 三层架构并行
|
|
```bash
|
|
# 表示层、业务层、数据层并行分析
|
|
(
|
|
gemini -p "@{src/views/**/*,src/components/**/*} @{CLAUDE.md} Presentation layer analysis" &
|
|
gemini -p "@{src/services/**/*,src/business/**/*} @{CLAUDE.md} Business logic layer analysis" &
|
|
gemini -p "@{src/models/**/*,src/db/**/*} @{CLAUDE.md} Data access layer analysis" &
|
|
wait
|
|
)
|
|
```
|
|
|
|
### 微服务架构并行
|
|
```bash
|
|
# 微服务并行分析
|
|
(
|
|
cd services/user-service && gemini --all-files -p "@{CLAUDE.md} User service patterns and architecture" &
|
|
cd services/order-service && gemini --all-files -p "@{CLAUDE.md} Order service patterns and architecture" &
|
|
cd services/payment-service && gemini --all-files -p "@{CLAUDE.md} Payment service patterns and architecture" &
|
|
cd services/notification-service && gemini --all-files -p "@{CLAUDE.md} Notification service patterns and architecture" &
|
|
wait
|
|
)
|
|
```
|
|
|
|
## 按功能领域并行
|
|
|
|
### 核心功能并行分析
|
|
```bash
|
|
# 核心业务功能并行分析
|
|
(
|
|
gemini -p "@{**/*auth*,**/*login*,**/*session*} @{CLAUDE.md} Authentication and session management analysis" &
|
|
gemini -p "@{**/api/**/*,**/routes/**/*,**/controllers/**/*} @{CLAUDE.md} API endpoints and routing analysis" &
|
|
gemini -p "@{**/components/**/*,**/ui/**/*,**/views/**/*} @{CLAUDE.md} UI components and interface analysis" &
|
|
gemini -p "@{**/models/**/*,**/entities/**/*,**/schemas/**/*} @{CLAUDE.md} Data models and schema analysis" &
|
|
wait
|
|
)
|
|
```
|
|
|
|
### 跨切面关注点并行
|
|
```bash
|
|
# 横切关注点并行分析
|
|
(
|
|
gemini -p "@{**/*security*,**/*crypto*,**/*auth*} @{CLAUDE.md} Security and encryption patterns analysis" &
|
|
gemini -p "@{**/*log*,**/*monitor*,**/*track*} @{CLAUDE.md} Logging and monitoring patterns analysis" &
|
|
gemini -p "@{**/*cache*,**/*redis*,**/*memory*} @{CLAUDE.md} Caching and performance patterns analysis" &
|
|
gemini -p "@{**/*test*,**/*spec*,**/*mock*} @{CLAUDE.md} Testing strategies and patterns analysis" &
|
|
wait
|
|
)
|
|
```
|
|
|
|
## 按技术栈并行
|
|
|
|
### 全栈技术并行分析
|
|
```bash
|
|
# 多技术栈并行分析
|
|
(
|
|
gemini -p "@{**/*.{js,jsx,ts,tsx}} @{CLAUDE.md} JavaScript/TypeScript patterns and usage analysis" &
|
|
gemini -p "@{**/*.{css,scss,sass,less}} @{CLAUDE.md} Styling patterns and CSS architecture analysis" &
|
|
gemini -p "@{**/*.{py,pyx}} @{CLAUDE.md} Python code patterns and implementation analysis" &
|
|
gemini -p "@{**/*.{sql,migration}} @{CLAUDE.md} Database schema and migration patterns analysis" &
|
|
wait
|
|
)
|
|
```
|
|
|
|
### 框架特定并行分析
|
|
```bash
|
|
# React 生态系统并行分析
|
|
(
|
|
gemini -p "@{src/components/**/*.{jsx,tsx}} @{CLAUDE.md} React component patterns and composition analysis" &
|
|
gemini -p "@{src/hooks/**/*.{js,ts}} @{CLAUDE.md} Custom hooks patterns and usage analysis" &
|
|
gemini -p "@{src/context/**/*.{js,ts,jsx,tsx}} @{CLAUDE.md} Context API usage and state management analysis" &
|
|
gemini -p "@{**/*.stories.{js,jsx,ts,tsx}} @{CLAUDE.md} Storybook stories and component documentation analysis" &
|
|
wait
|
|
)
|
|
```
|
|
|
|
## 按项目规模并行
|
|
|
|
### 大型项目分块并行
|
|
```bash
|
|
# 大型项目按模块并行分析
|
|
analyze_large_project() {
|
|
local modules=("auth" "user" "product" "order" "payment" "notification")
|
|
local pids=()
|
|
|
|
for module in "${modules[@]}"; do
|
|
(
|
|
echo "Analyzing module: $module"
|
|
gemini -p "@{src/$module/**/*,lib/$module/**/*} @{CLAUDE.md}
|
|
|
|
Module-specific analysis for $module:
|
|
1. Module architecture and organization patterns
|
|
2. Internal API and interface definitions
|
|
3. Integration points with other modules
|
|
4. Testing strategies and coverage
|
|
5. Performance considerations and optimizations
|
|
|
|
Focus on module-specific patterns and integration points."
|
|
) &
|
|
pids+=($!)
|
|
|
|
# 控制并行数量
|
|
if [ ${#pids[@]} -ge 3 ]; then
|
|
wait "${pids[0]}"
|
|
pids=("${pids[@]:1}") # 移除已完成的进程ID
|
|
fi
|
|
done
|
|
|
|
# 等待所有剩余进程
|
|
for pid in "${pids[@]}"; do
|
|
wait "$pid"
|
|
done
|
|
}
|
|
```
|
|
|
|
### 企业级项目并行策略
|
|
```bash
|
|
# 企业级项目分层并行分析
|
|
enterprise_parallel_analysis() {
|
|
# 第一层：核心架构分析
|
|
echo "Phase 1: Core Architecture Analysis"
|
|
(
|
|
gemini -p "@{src/core/**/*,lib/core/**/*} @{CLAUDE.md} Core architecture and foundation patterns" &
|
|
gemini -p "@{config/**/*,*.config.*} @{CLAUDE.md} Configuration management and environment setup" &
|
|
gemini -p "@{docs/**/*,README*,CHANGELOG*} @{CLAUDE.md} Documentation structure and project information" &
|
|
wait
|
|
)
|
|
|
|
# 第二层：业务模块分析
|
|
echo "Phase 2: Business Module Analysis"
|
|
(
|
|
gemini -p "@{src/modules/**/*} @{CLAUDE.md} Business modules and domain logic analysis" &
|
|
gemini -p "@{src/services/**/*} @{CLAUDE.md} Service layer and business services analysis" &
|
|
gemini -p "@{src/repositories/**/*} @{CLAUDE.md} Data access and repository patterns analysis" &
|
|
wait
|
|
)
|
|
|
|
# 第三层：基础设施分析
|
|
echo "Phase 3: Infrastructure Analysis"
|
|
(
|
|
gemini -p "@{infrastructure/**/*,deploy/**/*} @{CLAUDE.md} Infrastructure and deployment patterns" &
|
|
gemini -p "@{scripts/**/*,tools/**/*} @{CLAUDE.md} Build scripts and development tools analysis" &
|
|
gemini -p "@{tests/**/*,**/*.test.*} @{CLAUDE.md} Testing infrastructure and strategies analysis" &
|
|
wait
|
|
)
|
|
}
|
|
```
|
|
|
|
## 智能并行调度
|
|
|
|
### 依赖感知并行执行
|
|
```bash
|
|
# 基于依赖关系的智能并行调度
|
|
dependency_aware_parallel() {
|
|
local -A dependencies=(
|
|
["core"]=""
|
|
["utils"]="core"
|
|
["services"]="core,utils"
|
|
["api"]="services"
|
|
["ui"]="services"
|
|
["tests"]="api,ui"
|
|
)
|
|
|
|
local -A completed=()
|
|
local -A running=()
|
|
|
|
while [ ${#completed[@]} -lt ${#dependencies[@]} ]; do
|
|
for module in "${!dependencies[@]}"; do
|
|
# 跳过已完成或正在运行的模块
|
|
[[ ${completed[$module]} ]] && continue
|
|
[[ ${running[$module]} ]] && continue
|
|
|
|
# 检查依赖是否已完成
|
|
local deps="${dependencies[$module]}"
|
|
local can_start=true
|
|
|
|
if [[ -n "$deps" ]]; then
|
|
IFS=',' read -ra dep_array <<< "$deps"
|
|
for dep in "${dep_array[@]}"; do
|
|
[[ ! ${completed[$dep]} ]] && can_start=false && break
|
|
done
|
|
fi
|
|
|
|
# 启动模块分析
|
|
if $can_start; then
|
|
echo "Starting analysis for module: $module"
|
|
(
|
|
gemini -p "@{src/$module/**/*} @{CLAUDE.md} Module $module analysis"
|
|
echo "completed:$module"
|
|
) &
|
|
running[$module]=$!
|
|
fi
|
|
done
|
|
|
|
# 检查完成的进程
|
|
for module in "${!running[@]}"; do
|
|
if ! kill -0 "${running[$module]}" 2>/dev/null; then
|
|
completed[$module]=true
|
|
unset running[$module]
|
|
echo "Module $module analysis completed"
|
|
fi
|
|
done
|
|
|
|
sleep 1
|
|
done
|
|
}
|
|
```
|
|
|
|
### 资源自适应并行
|
|
```bash
|
|
# 基于系统资源的自适应并行
|
|
adaptive_parallel_execution() {
|
|
local available_memory=$(free -m 2>/dev/null | awk 'NR==2{print $7}' || echo 4000)
|
|
local cpu_cores=$(nproc 2>/dev/null || echo 4)
|
|
|
|
# 根据资源计算最优并行数
|
|
local max_parallel
|
|
if [ "$available_memory" -lt 2000 ]; then
|
|
max_parallel=2
|
|
elif [ "$available_memory" -lt 4000 ]; then
|
|
max_parallel=3
|
|
else
|
|
max_parallel=$((cpu_cores > 4 ? 4 : cpu_cores))
|
|
fi
|
|
|
|
echo "Adaptive parallel execution: $max_parallel concurrent processes"
|
|
|
|
local commands=(
|
|
"gemini -p '@{src/components/**/*} @{CLAUDE.md} Component analysis'"
|
|
"gemini -p '@{src/services/**/*} @{CLAUDE.md} Service analysis'"
|
|
"gemini -p '@{src/utils/**/*} @{CLAUDE.md} Utility analysis'"
|
|
"gemini -p '@{src/api/**/*} @{CLAUDE.md} API analysis'"
|
|
"gemini -p '@{src/models/**/*} @{CLAUDE.md} Model analysis'"
|
|
)
|
|
|
|
local active_jobs=0
|
|
for cmd in "${commands[@]}"; do
|
|
eval "$cmd" &
|
|
((active_jobs++))
|
|
|
|
# 达到并行限制时等待
|
|
if [ $active_jobs -ge $max_parallel ]; then
|
|
wait
|
|
active_jobs=0
|
|
fi
|
|
done
|
|
|
|
wait # 等待所有剩余任务完成
|
|
}
|
|
```
|
|
|
|
## 错误处理和监控
|
|
|
|
### 并行执行错误处理
|
|
```bash
|
|
# 带错误处理的并行执行
|
|
robust_parallel_execution() {
|
|
local commands=("$@")
|
|
local pids=()
|
|
local results=()
|
|
|
|
# 启动所有并行任务
|
|
for i in "${!commands[@]}"; do
|
|
(
|
|
echo "Starting task $i: ${commands[$i]}"
|
|
if eval "${commands[$i]}"; then
|
|
echo "SUCCESS:$i"
|
|
else
|
|
echo "FAILED:$i"
|
|
fi
|
|
) &
|
|
pids+=($!)
|
|
done
|
|
|
|
# 等待所有任务完成并收集结果
|
|
for i in "${!pids[@]}"; do
|
|
if wait "${pids[$i]}"; then
|
|
results+=("Task $i: SUCCESS")
|
|
else
|
|
results+=("Task $i: FAILED")
|
|
echo "Task $i failed, attempting retry..."
|
|
|
|
# 简单重试机制
|
|
if eval "${commands[$i]}"; then
|
|
results[-1]="Task $i: SUCCESS (retry)"
|
|
else
|
|
results[-1]="Task $i: FAILED (retry failed)"
|
|
fi
|
|
fi
|
|
done
|
|
|
|
# 输出执行结果摘要
|
|
echo "Parallel execution summary:"
|
|
for result in "${results[@]}"; do
|
|
echo " $result"
|
|
done
|
|
}
|
|
```
|
|
|
|
### 实时进度监控
|
|
```bash
|
|
# 带进度监控的并行执行
|
|
monitored_parallel_execution() {
|
|
local total_tasks=$#
|
|
local completed_tasks=0
|
|
local failed_tasks=0
|
|
|
|
echo "Starting $total_tasks parallel tasks..."
|
|
|
|
for cmd in "$@"; do
|
|
(
|
|
if eval "$cmd"; then
|
|
echo "COMPLETED:$(date): $cmd"
|
|
else
|
|
echo "FAILED:$(date): $cmd"
|
|
fi
|
|
) &
|
|
done
|
|
|
|
# 监控进度
|
|
while [ $completed_tasks -lt $total_tasks ]; do
|
|
sleep 5
|
|
|
|
# 计算当前完成数量
|
|
local current_completed=$(jobs -r | wc -l)
|
|
local current_failed=$((total_tasks - current_completed - $(jobs -s | wc -l)))
|
|
|
|
if [ $current_completed -ne $completed_tasks ] || [ $current_failed -ne $failed_tasks ]; then
|
|
completed_tasks=$current_completed
|
|
failed_tasks=$current_failed
|
|
|
|
echo "Progress: Completed: $completed_tasks, Failed: $failed_tasks, Remaining: $((total_tasks - completed_tasks - failed_tasks))"
|
|
fi
|
|
done
|
|
|
|
wait
|
|
echo "All parallel tasks completed."
|
|
}
|
|
```
|
|
|
|
这些并行执行模板提供了各种场景下的并行分析策略，从简单的并行执行到复杂的依赖感知调度和资源自适应执行。 |