01
Apex with AI
Where AI Apex generation works, where it breaks, and how to review the output without trusting it.
Resource HubTry Clientell AI free
Resources for Salesforce developers
Apex, LWC, and integration patterns for the engineers shipping production code on Salesforce.
Salesforce developers live inside a unique constraint envelope: governor limits, async patterns, metadata coupling, and a declarative-first culture that views code as a last resort. Generic AI coding advice doesn't translate. These resources are written by people who've shipped AI-generated Apex into real production orgs, and explain where it works, where it breaks, and how to keep it reviewable.
08 resources · 04 FAQs
“The most expensive Salesforce code is the Apex you didn't have to write. The second most expensive is the Apex you wrote because you didn't know AI could've.”
Top jobs to be done
What developers actually need help with.
02
LWC patterns
Component patterns that beat the AppExchange starter kits, with reactivity and state shape that scales.
03
Integration & API
Platform Events, callouts, async patterns. When to use which, and the governor-limit footguns.
04
Test class generation
Why AI-generated test classes hit 90% coverage but don't actually test anything, and the fix.
05
DevOps & deployment
Source-driven development, CI/CD on Salesforce, and the Copado/Gearset trade-offs.
06
Refactoring inherited Apex
The AI workflow for reading 8,000 lines of legacy Apex and figuring out what it actually does.
Use cases for developers
Explore by what you're trying to ship.
#01#02#03#04
AI flow builder
Generate bulk-safe Flow XML from natural language and review the diff before deployment. Faster than hand-writing for repetitive patterns.
Data operations + migration
Move data between sandbox and prod, migrate from legacy systems, run mass updates with rollback. AI plans the playbook.
Org documentation for inherited code
Auto-document the Apex you didn't write. Generate the architecture diagram, the dependency map, and the test-coverage gap report.
Data cleaning at scale
Run dedupe, validation, and standardization across multiple objects from a single agent. Better than DML scripts you maintain by hand.
By topic
Deep-dive playbooks for developers.
01 / Playbook
Developers Apex with AI
AI Apex generation is real, useful, and full of footguns. This page is the practitioner guide to using AI for Apex without produci…
02 / Playbook
Developers integration patterns
Integration is where Salesforce architecture choices show up first. The wrong pattern works fine for 100 records and dies at 100,0…
Curated for developers
Resources, hand-picked for this role.
Cheatsheet
/01
0/100
Score
The
63
Record-level
76
10
89
The
67
Permission Set Spaghetti: The Map Nobody Drew For You
The 4 permission layers and how they resolve, 10 named anti-patterns, a SOQL query bank and a cleanup sequence that will not break access.
Salesforce Security·PDF
Playbook
/02
0steps
Workflow
- 5-stage
- 02Mapped
- 03Includes
- Sandbox-safe
How to Fix Salesforce Flow Overlap (Race Conditions)
Step-by-step playbook to find and fix competing automations on Salesforce objects: detect race conditions, retire Workflow Rules, consolidate to one Flow.
Salesforce Automation
Playbook
/03
0steps
Dev workflow
- Generate Apex
- 02Read inherited code
- 03Generate tests
- Review with AI
The Salesforce Developer's AI Playbook
How working Salesforce developers are using AI to ship Apex, LWC, and integrations 3–5x faster, without giving up code review discipline or production safety.
Salesforce Development·14 min read
Template
/04
0/100
Score
A
83
A
61
Sandbox-first
74
One
87
CLAUDE.md Template for a Salesforce DX Project, with Skills
A CLAUDE.md and skills starter for a Salesforce DX project: org rules, sandbox-first habits and the commands to allow. Copy, edit, commit.
Salesforce AI
Template
/05
0/100
Score
Eight
66
What
79
A
92
Every
70
Where Is This Salesforce Field Used? Prompts for Coding Agents
Prompts that make Claude Code, Cursor, Codex or VS Code find every Flow, Apex trigger, validation rule and layout that uses a Salesforce field.
Salesforce AI
Template
/06
0/100
Score
100
87
Covers
65
3-hour
78
Works
91
100 AI Prompts That Do Your Salesforce Admin Job For You
100 copy-paste AI prompts across 10 admin categories: users, formulas, flows, debugging, reports, LWC, SOQL, Apex, data cleanup and documentation.
Salesforce AI·PDF
Checklist
/07
0/100
Score
The
77
Gate
90
The
68
The
81
Salesforce Record Access, Decoded: The Whole Sharing & Security Model on One Page
The Salesforce sharing and security model on one page: OWD, the 6-rung record-access ladder, restriction vs scoping rules, FLS and the Apex/Flow bug. Free PDF.
Salesforce Security·PDF
Checklist
/08
0/100
Score
15
65
Signal
78
Built
91
Vendor-neutra…
69
15 Salesforce Flows Quietly Tanking Your Org
A 15-minute audit checklist of the 15 most common Flow anti-patterns in production Salesforce orgs. Signal, why it hurts, and the rebuild for each.
Salesforce Flows·PDF
Adjacent resources
Things developers also look up.
AI Tool Directory · Developers26 tools · 6 jobs
Best AI tools for Salesforce developers.
Ranked directory with workflows, pricing, and pros and cons. Curated stack picks for developers below.
Starter stack for developers
See all →Frequently asked
Questions developers keep asking.
Should Salesforce developers worry about AI replacing them?
Not in the next 3 years, but the bar is rising fast. The developers who'll thrive are the ones who treat AI as a senior pair-programmer: review every line, push back on bad patterns, and own the architecture. The ones who treat it as autocomplete will be the first to go.
What's the biggest mistake devs make with AI-generated Apex?
Trusting the test classes. AI hits 90%+ coverage easily but the assertions are usually trivial (assertEquals on a value the test itself just set). You get a green build that proves nothing. The fix: write the assertion targets first, then ask AI to write the test scaffolding and DML setup around them.
Is Salesforce DX worth migrating to in 2026?
Yes, and the migration cost has dropped because AI can read your sandbox metadata and propose a source-driven structure for it. The pre-AI migration was a 3-week project. With Clientell AI or similar tooling it's 2 days. Stop deferring it.
How do I avoid governor-limit footguns when AI generates Apex?
Specify the bulk pattern in the prompt up front: 'Generate a trigger handler that bulk-safe processes up to 200 records per transaction, queries related records in a single SOQL with parent traversal, and DMLs the result set in one call.' AI will follow the constraint when you state it; it will produce non-bulk-safe code by default.
Getting Started
Skip the reading. Ship the work.
Resources only get you so far. Hand the actual developers work to Clientell AI, the agent that builds Flows, cleans data, and manages users on your real Salesforce org.
14-day free trial · No credit card required
SOC 2
HIPAA
ISO 27001
Salesforce Partner