Build Your Storefront
Learning Objectives
After completing this unit, you’ll be able to:
- Summarize the challenges teams face when using general-purpose AI coding tools for B2C Commerce development.
- Explain how the components of the Agentic B2C Developer Toolkit work together.
- Describe how coding agents use Agentic B2C Developer Toolkit Skills and MCP Servers to build and deploy Commerce apps without opening the Salesforce UI.
Build Custom Commerce Experiences
You ask an AI coding assistant to build a B2C Commerce feature. The generated code seems correct, but it uses the wrong APIs, misses platform-specific patterns, or doesn't follow best practices. Instead of saving time, you lose time from additional reviews, debugging, and code rewrites.
This challenge is common when using general-purpose AI coding tools. They understand programming languages but don't always understand the B2C Commerce platform.
In this unit, you explore how the Agentic B2C Developer Toolkit gives AI coding agents the B2C Commerce knowledge they need to generate accurate code, reduce rework, and help you spend more time building new features.
Meet the Agentic B2C Developer Toolkit
The Agentic B2C Developer Toolkit is one unified toolchain that works across every B2C storefront, and across every layer of the stack, from frontend to backend to operations. It gives developers and their coding agents everything they need to build, deploy, and operate B2C Commerce together using four components.
-
Command Line Interface (CLI): CLI is the foundation of the toolkit. Don't rely on outdated scripts that break at the worst times. Instead, use a consistent set of commands to deploy cartridges, run jobs, and manage your sandboxes, right from the terminal.
-
Agent Skills: Agent Skills are preconfigured playbooks that teach coding agents how B2C Commerce actually works and which CLI commands to run. You just give it the right playbook, and it starts generating accurate, production-ready code. Agent Skills work with the tools your team already uses such as Claude Code, Cursor, Codex, GitHub Copilot, Agentforce Vibes, and Windsurf.
-
Model Context Protocol (MCP) Server: A focused set of MCP tools that complement the CLI for agent-driven workflows. Writing code is just the beginning. The MCP Server gives your coding agent the tools to actually do things in your B2C Commerce environment such as deploy code, debug it live, and inspect what's happening under the hood. The MCP Server works together with Agent Skills, so your agent has both the knowledge of how B2C Commerce works and the tools to act on it.
-
Visual Studio (VS) Code Extension: Constantly switching between your IDE, Business Manager, and the terminal breaks your focus. The VS Code Extension keeps you in one place. It keeps you in the flow of work so you can review, debug, and approve the agent's code without constant context switching across systems.
Each component works independently, but the real advantage comes when they work together.
Bring the Components Together
Together, these four components fundamentally change how you create commerce applications. Agent Skills tell your coding agent how B2C Commerce works. The MCP Server gives it the tools to act. The CLI deploys and records the result. And the VS Code Extension is where you review, adjust, and approve the work.
For example, a developer needs to build a new Storefront Next storefront. Instead of manually wiring up backend services, they describe what they need and prompt their coding agent. The agent uses Agent Skills to understand the relevant APIs, writes the Storefront Next code, and uses the MCP Server and CLI to deploy bundles to a Managed Runtime environment for testing. The developer reviews the output in VS Code, debugs, and approves.
The result is a full development loop where the agent does the challenging tasks and the developer stays in control.
Explore the Toolkit in Action
When CLI, Agent Skills, and the MCP Server work together, they change how developers, merchandisers, and operators get things done everyday.
Triage Production Issues Faster
When order failures hit production, operators don't have to manually dig through logs or navigate Business Manager. A coding agent connected to the MCP Server can query the right APIs, surface the relevant diagnostics, and identify the root cause from the operator's existing workspace. The CLI handles any remediation such as redeploy a cartridge, roll back a job, or trigger a pipeline fix from a single command.
Investigate Site Problems
Bot attacks and performance degradation are complex to diagnose. With Agent Skills, a developer can ask their coding agent to run a bot investigation or performance diagnostic directly from their terminal or IDE. They don’t need to switch between monitoring tools.
Get Instant Site Analytics
Merchandisers can request reporting and analytics directly from a connected AI assistant, including promotion summaries and top-selling product reports. These tools provide immediate insights without the need to manually pull reports from the Business Manager.
Each scenario uses the same toolkit, just through a different lens depending on your role.
Wrap It Up
In this unit, you explored how the Agentic B2C Developer Toolkit helps teams build and operate headless B2C Commerce storefronts. You can now ship faster, debug smarter, and keep your team running smoothly as your coding agent does the high-effort work while you stay in control.