Back to Details

developer-toolbox

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Developer Toolbox

开发者工具箱

A collection of essential development workflow agents that integrate seamlessly with Claude Code.
这是一组可与Claude Code无缝集成的必备开发工作流Agent集合。

What's Included

包含内容

Agents (7)

Agent（7个）

AgentPurposeTriggers On
commit-helperGenerate conventional commit messages"commit message", "staged changes"
build-verifierVerify dist/ matches source after builds"changes not appearing", "verify build"
code-reviewerSecurity audits and code quality reviews"code review", "security audit", "OWASP"
debuggerSystematic debugging with root cause analysis"error", "TypeError", "stack trace", "bug"
test-runnerTDD workflow and test creation"write tests", "TDD", "coverage", "jest"
orchestratorCoordinate complex multi-step projects"coordinate", "multi-step", "complex feature"
documentation-expertREADME, API docs, architecture diagrams"document", "README", "API docs"
Agent用途触发条件
commit-helper生成符合规范的提交信息"commit message"、"staged changes"
build-verifier验证构建后dist/目录与源码是否匹配"changes not appearing"、"verify build"
code-reviewer安全审计与代码质量审查"code review"、"security audit"、"OWASP"
debugger系统化调试并进行根因分析"error"、"TypeError"、"stack trace"、"bug"
test-runner测试驱动开发（TDD）工作流与测试用例创建"write tests"、"TDD"、"coverage"、"jest"
orchestrator协调复杂的多步骤项目"coordinate"、"multi-step"、"complex feature"
documentation-expert编写README、API文档与架构图"document"、"README"、"API docs"

Rules (1)

规则（1个）

RulePurpose
agent-first-thinkingBehavioral interrupt - consider agents before manual work
规则用途
agent-first-thinking行为干预规则——优先考虑使用Agent而非手动操作

Installation

安装

bash
undefined
bash
undefined

Via marketplace

通过应用市场安装

/plugin install developer-toolbox
/plugin install developer-toolbox

Or local development

或本地开发环境安装

/plugin install ./skills/developer-toolbox

After installation, restart Claude Code to load the agents.
/plugin install ./skills/developer-toolbox

安装完成后，重启Claude Code以加载Agent。

Usage Examples

使用示例

Commit Helper

Commit Helper

"Help me write a commit message for these staged changes"
"帮我为这些暂存的变更编写提交信息"

Build Verifier

Build Verifier

"My changes aren't appearing in production, verify the build output"
"我的变更没有出现在生产环境中，请验证构建输出"

Code Reviewer

Code Reviewer

"Review this authentication code for security vulnerabilities"
"审查这段认证代码是否存在安全漏洞"

Debugger

Debugger

"I'm getting TypeError: Cannot read property 'map' of undefined"
"我遇到了TypeError: Cannot read property 'map' of undefined错误"

Test Runner

Test Runner

"Use TDD to implement this user validation function"
"使用TDD方式实现这个用户验证函数"

Orchestrator

Orchestrator

"Coordinate a refactor of the authentication system across 5 services"
"协调跨5个服务的认证系统重构工作"

Documentation Expert

Documentation Expert

"Create comprehensive API documentation for this REST endpoint"
"为这个REST端点创建全面的API文档"

Agent Design Philosophy

Agent设计理念

All agents follow the "MUST BE USED when" pattern for reliable auto-discovery:
yaml
description: |
  [Role] specialist. MUST BE USED when: [trigger 1], [trigger 2], [trigger 3].
  Use PROACTIVELY for [broad task category].

  Keywords: keyword1, keyword2, error-message-fragment
This ensures Claude Code discovers and proposes the right agent automatically based on user requests.
所有Agent均遵循**"必须在以下场景使用"**的模式，以确保可靠的自动发现：
yaml
description: |
  [角色] 专家。必须在以下场景使用：[触发场景1]、[触发场景2]、[触发场景3]。
  主动用于[广泛任务类别]。

  关键词：keyword1, keyword2, error-message-fragment
这确保Claude Code能根据用户请求自动发现并推荐合适的Agent。

Agent-First Thinking Rule

Agent优先思维规则

The included 
agent-first-thinking.md
rule encourages using agents by default:
The Inversion:
  • Wrong: "I'll do this manually unless it's big enough for agents"
  • Right: "I'll use agents unless there's a reason not to"
Triggers:
If about to...Use instead...
grep/glob 3+ timesExplore agent
Read 5+ filesExplore agent
Same edit across filesParallel agents
Audit multiple itemsParallel swarm
内置的
agent-first-thinking.md
规则鼓励默认使用Agent：
思维转变：
  • 错误做法："除非任务足够大，否则我手动完成"
  • 正确做法："除非有特殊理由，否则我使用Agent"
触发场景：
当你要...替代方案...
执行3次以上grep/glob操作尝试使用Agent
阅读5个以上文件尝试使用Agent
在多个文件中执行相同编辑使用并行Agent
审计多个项目项使用并行Agent集群

Customization

自定义

Each agent can be extended by editing its markdown file after installation:
bash
undefined
安装完成后，可通过编辑每个Agent的markdown文件进行扩展：
bash
undefined

Find installed agents

查找已安装的Agent

ls ~/.claude/plugins/cache//developer-toolbox//agents/
ls ~/.claude/plugins/cache//developer-toolbox//agents/

Or copy to user-level for customization

或复制到用户目录进行自定义

cp [plugin-path]/agents/code-reviewer.md ~/.claude/agents/
undefined
cp [plugin-path]/agents/code-reviewer.md ~/.claude/agents/
undefined

Combining Agents

Agent组合使用

Agents work well together:
"Review this code for security issues, then write tests for the critical paths"
Agent可以协同工作：
"审查这段代码的安全问题，然后为关键路径编写测试用例"

→ code-reviewer first, then test-runner

→ 先调用code-reviewer，再调用test-runner

undefined
"Debug this failing test, document the root cause, and commit the fix"
undefined
"调试这个失败的测试用例，记录根因分析结果，并提交修复代码"

→ debugger → documentation-expert → commit-helper

→ debugger → documentation-expert → commit-helper

undefined
undefined

Version History

版本历史

  • 1.0.0 (2025-01-20): Initial release with 7 agents and 1 rule
  • 1.0.0（2025-01-20）：初始版本，包含7个Agent和1条规则