Claude Code实战 Salesforce开发 MCP智能代理工作流

这是一门面向Salesforce开发者/架构师的实战课程 教你如何使用Claude Code进行AI驱动的Salesforce开发 从基础到构建完整的智能代理团队

课程时长：5小时 1920X1080 mp4 语言：英语+Ai中英文字幕 含课程文件

课程核心内容：
• Claude Code基础设置与终端工作流
• Apex开发（含批量Apex）验证规则 记录触发流程
• 从图片原型生成Lightning Web组件
• MCP企业级集成（Salesforce Jira GitHub联动）
• 多智能代理系统架构（设计 开发 测试 运维7大代理）
• 上下文管理与令牌优化策略
• 自定义命令与安全权限控制

适合人群：
Salesforce开发者/管理员/技术架构师
AI应用探索者
DevOps工程师

学习收获：
掌握Claude Code企业级应用
打通Salesforce GitHub Jira全流程
构建AI智能代理团队
交付真实用户故事

课程要求：
了解Salesforce基础
能运行SF CLI和Claude Code的电脑
无需Claude Code/MCP经验

#Salesforce #ClaudeCode #AI编程 #MCP协议 #开发课程







Claude Code for Salesforce Development: MCP & Agent Workflow

Master Claude Code, MCP, and Agent Workflows for AI-Driven Salesforce Development from Scratch to Create Your Agent Team

Claude Code for Salesforce Development: MCP & Agent Workflows

Salesforce development is evolving.

This course teaches you how to move beyond basic prompting and start building production-grade AI workflows inside real Salesforce projects using Claude Code.

This is a complete, end-to-end guide to mastering Claude Code for Salesforce development, from fundamentals to building a coordinated Salesforce Agent Team that can deliver a full user story.

If you are a Salesforce Developer, Salesforce Admin, Technical Architect, or AI enthusiast working in the Salesforce ecosystem, this course will show you how to leverage Claude Code, MCP (Model Context Protocol), and multi-agent systems to build faster, smarter, and more structured workflows.

What You Will Learn

Foundation and Setup

• What Claude Code is and how it works inside a terminal-first workflow

• Step-by-step setup of Claude Code for Salesforce DX projects

• Essential slash commands and custom commands

• Efficient context management and token optimization

• Plan Mode vs Edit Mode and structured AI reasoning

Core Salesforce Development with Claude Code

• Building Apex classes including Batch Apex

• Creating Validation Rules and Record-Triggered Flows

• Managing CLAUDE. md memory for persistent project context

• Using image mockups to generate Lightning Web Components

• Understanding Context, Tokens, Scope, and best practices for LLM-driven development

• Checkpoints and rollback strategies for safe AI-assisted coding

• Permission management using settings json for secure automation

Advanced AI Workflows and Productivity

• Building reusable custom slash commands

• Running Claude safely with controlled read and write permissions

• Designing structured AI workflows instead of ad-hoc prompts

• Optimizing token usage and reducing unnecessary context consumption

MCP (Model Context Protocol) and Enterprise Integrations

• Introduction to MCP architecture

• Enabling Salesforce MCP for intelligent deployments

• Connecting Claude Code with Salesforce, Jira, and GitHub

• Reading Jira user stories directly from Claude

• Creating Git branches and raising Pull Requests automatically

• Executing end-to-end workflows across multiple enterprise systems

This course teaches real-world DevOps workflows integrated with Salesforce DX, GitHub repositories, and Jira-based agile processes.

Multi-Agent Systems and Salesforce Agent Teams

• Understanding Sub Agents in Claude Code

• Creating specialized Salesforce agents (Design, Admin, Dev)

• Defining workflows using claude . md

• Scaling from a core team to a 7-Agent Salesforce architecture

• Coordinating Design, Development, Testing, Code Review, DevOps, and Documentation agents

• Delivering a complete Salesforce user story using shared memory and structured collaboration

You will learn how to architect and orchestrate AI-driven Salesforce teams using modern agent-based workflows.

Why This Course Is Different

This is not just another AI tutorial.

It focuses specifically on

• Salesforce DX development

• Apex and Flow automation

• Lightning Web Components

• Enterprise Git workflows

• Agile Jira integration

• AI agent architecture

• Context engineering

• Token-efficient AI workflows

By the end of this course, you will be able to

• Use Claude Code professionally inside Salesforce projects

• Connect Salesforce, GitHub, and Jira using MCP

• Optimize context and token usage

• Build reusable slash commands

• Implement safe automation permissions

• Design and deploy multi-agent Salesforce systems

• Deliver real user stories using AI coordination

What you’ll learn
✓ Master Claude Code for Salesforce Development from fundamentals to advanced agent workflows
✓ Understand and optimize context, tokens, and scope for efficient AI-driven Salesforce development
✓ Build Apex classes, Validation Rules, Flows, and LWCs using Claude Code
✓ Generate Lightning Web Components from image mockups using AI
✓ Manage project memory effectively using CLAUDE. md for structured development
✓ Create and use custom slash commands to improve productivity and reusability
✓ Implement safe automation using permissions and settings . json in Claude Code
✓ Use checkpoints and rollback features for controlled AI-assisted development
✓ Integrate Salesforce, Jira, and GitHub using MCP for end-to-end workflows
✓ Read Jira stories, deploy to Salesforce, create branches, and raise pull requests using Claude
✓ Design and build specialized Salesforce Sub Agents in Claude Code
✓ Create and orchestrate a multi-agent Salesforce team to deliver a complete user story

Requirements
● Basic understanding of Salesforce platform
● A computer capable of running Salesforce CLI and Claude Code
● No prior experience with Claude Code or MCP is required