Claude子代理创建指南：创建自动化子代理提升AI开发效率

create-subagents by glittercowboy/taches-cc-resources

101 周安装量

1,700 GitHub Stars

GitHub

安装命令

npx skills add https://github.com/glittercowboy/taches-cc-resources --skill create-subagents

AI/机器学习自动化代码质量

🇨🇳中文介绍

子代理可将复杂任务委托给专门的代理，这些代理无需用户交互即可自主运行，并将其最终输出返回至主对话。

<quick_start>

运行 /agents 命令
选择“创建新代理”
选择项目级别（.claude/agents/）或用户级别（~/.claude/agents/）
定义子代理：
- name : 小写字母加连字符
- description : 应在何时使用此子代理？
- tools : 可选的逗号分隔列表（若省略则继承所有工具）
- model : 可选（sonnet、opus、haiku 或 inherit）
编写系统提示（子代理的指令）

<configuration> <field name="name"> - 仅限小写字母和连字符 - 必须唯一 </field> <field name="description"> - 用途的自然语言描述 - 包含 Claude 应在何时调用此子代理 - 用于自动选择子代理 </field> <field name="tools"> - 逗号分隔列表：`Read, Write, Edit, Bash, Grep` - 如果省略：继承主线程的所有工具 - 使用 `/agents` 界面查看所有可用工具 </field> <field name="model"> - `sonnet`、`opus`、`haiku` 或 `inherit` - `inherit`：使用与主对话相同的模型 - 如果省略：默认为配置的子代理模型（通常为 sonnet） </field> </configuration> <principle name="use_pure_xml_structure"> 使用纯 XML 标签构建系统提示。从正文中移除所有 Markdown 标题。

广告位招租

在这里展示您的产品或服务

触达数万 AI 开发者，精准高效

联系我们

相关 Skills

FlyClaw：零登录航班聚合查询工具，Python实现多源航班信息与价格搜索

4,000,000 周安装

find-skills 技能搜索工具 - Vercel Labs 开源智能体技能包管理器

867,400 周安装

Azure RBAC 权限管理工具：查找最小角色、创建自定义角色与自动化分配

131,500 周安装

React 组合模式指南：Vercel 组件架构最佳实践，提升代码可维护性

116,600 周安装

代码质量和可维护性
安全漏洞
性能问题
最佳实践遵循情况</focus_areas>

<output_format> 提供具体、可操作的反馈，并附上文件:行号引用。</output_format>

</example>
</quick_start>

类型	位置	作用范围	优先级
项目	`.claude/agents/`	仅当前项目	最高
用户	`~/.claude/agents/`	所有项目	较低
插件	插件的 `agents/` 目录	所有项目	最低

当名称冲突时，项目级别子代理会覆盖用户级别子代理。 </file_structure>

<execution_model> <critical_constraint> 子代理是无法与用户交互的黑盒。

子代理在隔离的上下文中运行，并将其最终输出返回至主对话。它们：

✅ 可以使用如 Read、Write、Edit、Bash、Grep、Glob 等工具
✅ 可以访问 MCP 服务器和其他非交互式工具
❌ 不能使用 AskUserQuestion 或任何需要用户交互的工具
❌ 不能提供选项或等待用户输入
❌ 用户永远看不到子代理的中间步骤

主对话只能看到子代理的最终报告/输出。 </critical_constraint>

<workflow_design> 设计使用子代理的工作流：

使用主聊天进行：

从用户处收集需求（AskUserQuestion）
向用户呈现选项或决策
任何需要用户确认/输入的任务
用户需要了解进度的任务

使用子代理进行：

研究任务（API 文档查找、代码分析）
基于预定义需求的代码生成
分析和报告（安全审查、测试覆盖率）
不需要用户交互的上下文密集型操作

示例工作流模式：

主聊天：向用户询问需求（AskUserQuestion）↓ 子代理：研究 API 并创建文档（无需用户交互）↓ 主聊天：与用户一起审查研究，确认方法↓ 子代理：基于确认的计划生成代码↓ 主聊天：呈现结果，处理测试/部署

</workflow_design> </execution_model>

<system_prompt_guidelines> <principle name="be_specific"> 明确定义子代理的角色、能力和约束。 </principle>

---
name: security-reviewer
description: Reviews code for security vulnerabilities
tools: Read, Grep, Glob, Bash
model: sonnet
---

<role>
You are a senior code reviewer specializing in security.
</role>

<focus_areas>
- SQL injection vulnerabilities
- XSS attack vectors
- Authentication/authorization issues
- Sensitive data exposure
</focus_areas>

<workflow>
1. Read the modified files
2. Identify security risks
3. Provide specific remediation steps
4. Rate severity (Critical/High/Medium/Low)
</workflow>

❌ 不好："You are a helpful assistant that helps with code" ✅ 好："You are a React component refactoring specialist. Analyze components for hooks best practices, performance anti-patterns, and accessibility issues." </system_prompt_guidelines>

<subagent_xml_structure> 子代理 .md 文件是仅供 Claude 使用的系统提示。与技能和斜杠命令一样，它们应使用纯 XML 结构以实现最佳解析和令牌效率。

<recommended_tags> 子代理结构的常用标签：

<role> - 子代理是谁以及它做什么
<constraints> - 硬性规则（NEVER/MUST/ALWAYS）
<focus_areas> - 优先处理的事项
<workflow> - 分步流程
<output_format> - 如何构建交付物
<success_criteria> - 完成标准
<validation> - 如何验证工作 </recommended_tags>

<intelligence_rules> 简单子代理（单一专注任务）：

至少使用 role + constraints + workflow
示例：code-reviewer、test-runner

中等子代理（多步骤流程）：

添加 workflow steps、output_format、success_criteria
示例：api-researcher、documentation-generator

复杂子代理（研究 + 生成 + 验证）：

酌情添加所有标签，包括 validation、examples
示例：mcp-api-researcher、comprehensive-auditor </intelligence_rules>

<critical_rule> 从子代理正文中移除所有 Markdown 标题（##、###）。 改用语义化的 XML 标签。

在内容内部保留 Markdown 格式（粗体、斜体、列表、代码块、链接）。

有关 XML 结构原则和令牌效率的详细信息，请参阅 @skills/create-agent-skills/references/use-xml-tags.md - 同样的原则适用于子代理。 </critical_rule> </subagent_xml_structure>

使用 code-reviewer 子代理检查我最近的更改

让 test-writer 子代理为新 API 端点创建测试

<manual_editing> 您也可以直接编辑子代理文件：

项目：.claude/agents/subagent-name.md
用户：~/.claude/agents/subagent-name.md </manual_editing>

子代理使用和配置：references/subagents.md

文件格式和配置
模型选择（Sonnet 4.5 + Haiku 4.5 编排）
工具安全性和最小权限原则
提示缓存优化
完整示例

编写有效的提示：references/writing-subagent-prompts.md

核心原则和 XML 结构
用于路由的描述字段优化
复杂推理的扩展思考
安全约束和强模态动词
成功标准定义

评估指标（任务完成度、工具正确性、鲁棒性）
测试策略（离线、模拟、在线监控）
评估驱动开发
用于自定义标准的 G-Eval

错误处理和恢复：references/error-handling-and-recovery.md

常见故障模式和原因
恢复策略（优雅降级、重试、断路器）
结构化通信和可观测性
应避免的反模式

上下文管理：references/context-management.md

内存架构（STM、LTM、工作内存）
上下文策略（摘要、滑动窗口、草稿纸）
管理长时间运行的任务
提示缓存交互

顺序、并行、分层、协调器模式
用于成本/性能的 Sonnet + Haiku 编排
多代理协调
模式选择指南

调试和故障排除：references/debugging-agents.md

日志记录、跟踪和关联 ID
常见故障类型（幻觉、格式错误、工具误用）
诊断程序
持续监控

<success_criteria> 一个配置良好的子代理应具备：

有效的 YAML 前置元数据（名称与文件匹配，描述包含触发条件）
系统提示中清晰的角色定义
适当的工具限制（最小权限）
采用 XML 结构的系统提示，包含角色、方法和约束
为自动路由优化的描述字段
在代表性任务上成功测试
适合任务复杂度的模型选择（Sonnet 用于推理，Haiku 用于简单任务） </success_criteria>

🇺🇸English

Subagents enable delegation of complex tasks to specialized agents that operate autonomously without user interaction, returning their final output to the main conversation.

<quick_start>

Run /agents command
Select "Create New Agent"
Choose project-level (.claude/agents/) or user-level (~/.claude/agents/)
Define the subagent:
- name : lowercase-with-hyphens
- description : When should this subagent be used?
- tools : Optional comma-separated list (inherits all if omitted)
- model : Optional (sonnet, opus, haiku, or inherit)
Write the system prompt (the subagent's instructions)

<focus_areas>

Code quality and maintainability
Security vulnerabilities
Performance issues
Best practices adherence </focus_areas>

<output_format> Provide specific, actionable feedback with file:line references. </output_format>

</example>
</quick_start>

<file_structure>
| Type | Location | Scope | Priority |
|------|----------|-------|----------|
| **Project** | `.claude/agents/` | Current project only | Highest |
| **User** | `~/.claude/agents/` | All projects | Lower |
| **Plugin** | Plugin's `agents/` dir | All projects | Lowest |

Project-level subagents override user-level when names conflict.
</file_structure>

<configuration>
<field name="name">
- Lowercase letters and hyphens only
- Must be unique
</field>

<field name="description">
- Natural language description of purpose
- Include when Claude should invoke this subagent
- Used for automatic subagent selection
</field>

<field name="tools">
- Comma-separated list: `Read, Write, Edit, Bash, Grep`
- If omitted: inherits all tools from main thread
- Use `/agents` interface to see all available tools
</field>

<field name="model">
- `sonnet`, `opus`, `haiku`, or `inherit`
- `inherit`: uses same model as main conversation
- If omitted: defaults to configured subagent model (usually sonnet)
</field>
</configuration>

<execution_model>
<critical_constraint>
**Subagents are black boxes that cannot interact with users.**

Subagents run in isolated contexts and return their final output to the main conversation. They:
- ✅ Can use tools like Read, Write, Edit, Bash, Grep, Glob
- ✅ Can access MCP servers and other non-interactive tools
- ❌ **Cannot use AskUserQuestion** or any tool requiring user interaction
- ❌ **Cannot present options or wait for user input**
- ❌ **User never sees subagent's intermediate steps**

The main conversation sees only the subagent's final report/output.
</critical_constraint>

<workflow_design>
**Designing workflows with subagents:**

Use **main chat** for:
- Gathering requirements from user (AskUserQuestion)
- Presenting options or decisions to user
- Any task requiring user confirmation/input
- Work where user needs visibility into progress

Use **subagents** for:
- Research tasks (API documentation lookup, code analysis)
- Code generation based on pre-defined requirements
- Analysis and reporting (security review, test coverage)
- Context-heavy operations that don't need user interaction

**Example workflow pattern:**

Main Chat: Ask user for requirements (AskUserQuestion) ↓ Subagent: Research API and create documentation (no user interaction) ↓ Main Chat: Review research with user, confirm approach ↓ Subagent: Generate code based on confirmed plan ↓ Main Chat: Present results, handle testing/deployment

</workflow_design>
</execution_model>

<system_prompt_guidelines>
<principle name="be_specific">
Clearly define the subagent's role, capabilities, and constraints.
</principle>

<principle name="use_pure_xml_structure">
Structure the system prompt with pure XML tags. Remove ALL markdown headings from the body.

```markdown
---
name: security-reviewer
description: Reviews code for security vulnerabilities
tools: Read, Grep, Glob, Bash
model: sonnet
---

<role>
You are a senior code reviewer specializing in security.
</role>

<focus_areas>
- SQL injection vulnerabilities
- XSS attack vectors
- Authentication/authorization issues
- Sensitive data exposure
</focus_areas>

<workflow>
1. Read the modified files
2. Identify security risks
3. Provide specific remediation steps
4. Rate severity (Critical/High/Medium/Low)
</workflow>

❌ Bad: "You are a helpful assistant that helps with code" ✅ Good: "You are a React component refactoring specialist. Analyze components for hooks best practices, performance anti-patterns, and accessibility issues." </system_prompt_guidelines>

<subagent_xml_structure> Subagent.md files are system prompts consumed only by Claude. Like skills and slash commands, they should use pure XML structure for optimal parsing and token efficiency.

<recommended_tags> Common tags for subagent structure:

<role> - Who the subagent is and what it does
<constraints> - Hard rules (NEVER/MUST/ALWAYS)
<focus_areas> - What to prioritize
<workflow> - Step-by-step process
<output_format> - How to structure deliverables
<success_criteria> - Completion criteria
<validation> - How to verify work </recommended_tags>

<intelligence_rules> Simple subagents (single focused task):

Use role + constraints + workflow minimum
Example: code-reviewer, test-runner

Medium subagents (multi-step process):

Add workflow steps, output_format, success_criteria
Example: api-researcher, documentation-generator

Complex subagents (research + generation + validation):

Add all tags as appropriate including validation, examples
Example: mcp-api-researcher, comprehensive-auditor </intelligence_rules>

<critical_rule> Remove ALL markdown headings (##, ###) from subagent body. Use semantic XML tags instead.

Keep markdown formatting WITHIN content (bold, italic, lists, code blocks, links).

For XML structure principles and token efficiency details, see @skills/create-agent-skills/references/use-xml-tags.md - the same principles apply to subagents. </critical_rule> </subagent_xml_structure>

> Use the code-reviewer subagent to check my recent changes



> Have the test-writer subagent create tests for the new API endpoints

<manual_editing> You can also edit subagent files directly:

Project: .claude/agents/subagent-name.md
User: ~/.claude/agents/subagent-name.md </manual_editing>

Subagent usage and configuration : references/subagents.md

File format and configuration
Model selection (Sonnet 4.5 + Haiku 4.5 orchestration)
Tool security and least privilege
Prompt caching optimization
Complete examples

Writing effective prompts : references/writing-subagent-prompts.md

Core principles and XML structure
Description field optimization for routing
Extended thinking for complex reasoning
Security constraints and strong modal verbs
Success criteria definition

Advanced topics :

Evaluation and testing : references/evaluation-and-testing.md

Evaluation metrics (task completion, tool correctness, robustness)
Testing strategies (offline, simulation, online monitoring)
Evaluation-driven development
G-Eval for custom criteria

Error handling and recovery : references/error-handling-and-recovery.md

Common failure modes and causes
Recovery strategies (graceful degradation, retry, circuit breakers)
Structured communication and observability
Anti-patterns to avoid

Context management : references/context-management.md

Memory architecture (STM, LTM, working memory)
Context strategies (summarization, sliding window, scratchpads)
Managing long-running tasks
Prompt caching interaction

Orchestration patterns : references/orchestration-patterns.md

Sequential, parallel, hierarchical, coordinator patterns
Sonnet + Haiku orchestration for cost/performance
Multi-agent coordination
Pattern selection guidance

Debugging and troubleshooting : references/debugging-agents.md

Logging, tracing, and correlation IDs
Common failure types (hallucinations, format errors, tool misuse)
Diagnostic procedures
Continuous monitoring

<success_criteria> A well-configured subagent has:

Valid YAML frontmatter (name matches file, description includes triggers)
Clear role definition in system prompt
Appropriate tool restrictions (least privilege)
XML-structured system prompt with role, approach, and constraints
Description field optimized for automatic routing
Successfully tested on representative tasks
Model selection appropriate for task complexity (Sonnet for reasoning, Haiku for simple tasks) </success_criteria>

Weekly Installs

101

Repository

glittercowboy/t…esources

GitHub Stars

1.7K

First Seen

Jan 23, 2026

Security Audits

Gen Agent Trust HubPass SocketPass SnykWarn

Installed on

opencode85

codex82

claude-code78

gemini-cli78

cursor73

github-copilot73

AI Elements：基于shadcn/ui的AI原生应用组件库，快速构建对话界面

66,200 周安装

Claude子代理创建指南：创建自动化子代理提升AI开发效率

🇨🇳中文介绍

相关 Skills

🇺🇸English

最新 Skills