# Stack Salesforce Apex

> Reviews Salesforce Apex changes when governor limits, bulk processing, triggers, sharing, or tests need focused checks; keeps deployment and org writes separately authorized.

- Skill: `jovanipink/stack-salesforce-apex` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx get-skill add jovanipink/stack-salesforce-apex`
- Raw SKILL.md: https://api.skillmd.com/api/skills/jovanipink/stack-salesforce-apex/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: MIT
- Author: JovaniPink (https://skillmd.com/u/jovanipink)
- Updated: 2026-11-15
- Page: https://skillmd.com/skills/jovanipink/stack-salesforce-apex

---


# Salesforce, Apex, and LWC Engineering Profile

Use this profile only after repository discovery identifies the stack. Read [focused checks](references/checks.md) when Salesforce-specific gates or hazards determine the result.

## Workflow

1. Discover project configuration, package directories and dependencies, API versions, metadata manifests, org-shape assumptions, Apex, Lightning Web Components, permission metadata, destructive-change manifests, and repository scripts.
2. Follow repository-defined commands and pinned Salesforce CLI, Node, and package versions before suggesting defaults.
3. Review Apex hazards, especially bulk behavior, governor limits, record sharing, object and field permissions, query and DML placement, transaction boundaries, callouts, asynchronous work, recursion, and test data isolation.
4. Review Lightning hazards, including client-server contracts, reactive state, wire and cache behavior, browser security, accessibility, and permission-aware data access.
5. Select proportionate local static, lint, unit, package, manifest, and build gates. Treat any org-backed validation as external state requiring exact target-org and credential authority.
6. Review compatibility across API versions, package dependencies, namespaces, org features, metadata availability, Apex contracts, LWC targets, permissions, and deployment order.
7. Report every missing tool, skipped command, unavailable org, or unverified feature as incomplete rather than passing.

## Boundaries

- Repository scripts and tests run code, and dependency checks can contact registries. Run a gate only when the user asked for checks and trusts the repository. Otherwise, name it as the next safe gate.
- Do not infer a default org, use credentials, create a scratch org, install a package, run an org-backed test, deploy, or apply destructive changes without separate authority.
- A local parse, lint, or generated deployment archive does not prove org acceptance or live behavior.
- Keep metadata review, local validation, org validation, deployment, permission assignment, and live verification as separate evidence states.

## Output

Lead with the answer. Use the shortest format that retains the required evidence and limits; expand when requested or needed. The items below describe required content, not mandatory headings.

Return Discovery, Hazards, Commands selected, Results, Compatibility, Org evidence boundary, Missing evidence, and Next safe gate.

## Example

Synthetic example; not an observed client result.

Input: A trigger performs one query per record.

Output: warning: query count grows with batch size. Review bulk behavior and governor limits; deployment to an org is outside this review.

