Workshops: Agentic Coding by Tech Stack

Salesforce Development with AI Agents: Apex & LWC Workshop

Salesforce orgs hold a lot of business logic, much of it in Apex, Lightning Web Components and metadata that agents can read like any other source. What agents tend to miss are the platform's own rules: governor limits, bulkification, sharing and field-level security. This workshop teaches Salesforce developers to build features with coding agents in a source-driven SFDX workflow that respects those rules.

Who is this workshop for?

  • Salesforce developers working with Apex, LWC and Salesforce CLI
  • Technical architects and release managers in Salesforce teams
  • Experienced admins who increasingly write code
  • Consulting partners delivering Salesforce projects for several clients
  • ISV teams preparing AppExchange packages and the security review

What you will take away

  • Run agents in an SFDX project with scratch orgs or sandboxes, never directly against production.
  • Get bulkified Apex from agents that respects governor limits, plus tests that check behaviour instead of just producing coverage.
  • Build and test Lightning Web Components with Jest together with the agent.
  • Catch SOQL in loops, missing access checks and hard-coded IDs with Salesforce Code Analyzer and review.
  • Decide as a team when a requirement belongs in Flow, Apex or an LWC, and document it for the agent.

Agenda

  1. Module 1

    Salesforce projects for agents

    • Retrieving metadata into a Git repository with Salesforce CLI
    • Claude Code, Cursor, GitHub Copilot and Salesforce's own AI tooling compared
    • Context files: object model, naming conventions, org-specific rules
  2. Module 2

    Apex done right

    • Exercise: trigger handler with bulk-safe logic and tests
    • Governor limits, SOQL and DML outside loops, asynchronous Apex
    • Sharing, object and field permissions, user mode queries
  3. Module 3

    Lightning Web Components

    • Generating components with wire adapters and Apex controllers
    • Jest tests for LWC
    • Accessibility and the Salesforce Lightning Design System
  4. Module 4

    Testing and quality gates

    • Test data factories instead of SeeAllData
    • Salesforce Code Analyzer as feedback for the agent
    • Validation deployments and test runs in CI
  5. Module 5

    Release and context

    • Scratch orgs, sandboxes and the path to production
    • Agentforce in context: AI for end users versus AI help for developers

Platform rules agents need to learn

On Salesforce, code that works for one record can fail for two hundred. Agents with general programming knowledge readily put queries inside loops, ignore sharing rules or hard-code record IDs. Salesforce also requires at least 75 percent Apex test coverage for production deployments, which tempts agents to write tests that cover lines without checking behaviour.

We turn these rules into project guidelines, review checklists and automated checks, so the agent gets feedback before a deployment fails.

  • Bulkify every trigger and service method
  • Enforce sharing and field-level security
  • No hard-coded IDs or credentials
  • Assertions in every test, not just coverage

Org data and confidentiality

Agents work on metadata and code, not customer records. We use scratch orgs or developer sandboxes with test data and show how to keep personal data from production orgs out of prompts and logs, in line with GDPR. Which AI provider may process your source code is prepared together with your IT and data protection officers; the legal assessment stays with them.

Bring a representative part of your org's metadata or a sandbox. The workshop runs remotely or on-site, in English or German, with neooutput AI Code Engineers.

Frequently asked questions

Do we need Agentforce to take part?

No. The focus is on coding agents that work on your source code, independent of Salesforce licences. We put Agentforce and Salesforce's own developer tooling in context so you know where they complement general agents.

May an agent deploy to our org?

To scratch orgs and developer sandboxes, yes, if you allow it. Production deployments should go through your release process with validation, tests and human approval.

Is this suitable for admins who mainly build flows?

Partly. The workshop is code-focused. Admins with some Apex experience benefit most from the Flow-versus-code module; pure low-code teams are better served by a tailored format.

Workshop

Get in touch

Want agents to write Apex that survives governor limits and code review?

Tell us roughly how many people will join and pick dates that work – we will get back to you within 24 hours with a proposal for content, format and schedule.

Prefer to talk first?

15 minutes, free and without obligation – with our AI Trainers and AI Code Engineers.

Book a call for free (opens in new tab) +49 761 519 751 50 (24/7)

Request a training / workshop

Workshop: Salesforce Development with AI Agents: Apex & LWC Workshop · 1 day

Dates that work for you (optional)

Suggestions about 2 weeks ahead – pick as many as you like

Reply usually within 24 hours · jocapps is ISO 27001 certified · Privacy