MeshMeshBeta
Recipes

Security Audit for Salesforce Platform

A read-only security audit that shows what to fix first in your Salesforce org.

Overview

A read-only security audit of one connected Salesforce org, for admins, developers, architects and executives. It answers one question: where is your org exposed, and what should you fix first?

Nothing in your org changes. Every number comes from this run, and names and emails are masked throughout.

What you get

Each run produces three outputs:

  • Interactive report canvas — opens with your org facts, an overall score, severity counts, a five-sentence brief and up to five priority actions
  • Evidence file — every check with a stable rule ID, a control register and a score for each of five areas, with the formula shown
  • Fix guidance — every serious finding gets a Setup menu path or a code location, what the change might affect and how to confirm it's done

Inside the report

  • Five areas — identity and access; session and login, including logging and monitoring; integrations; data exposure and guest access; and automation and code
  • Score by area — stacked bars by severity. Click a segment to see the findings behind it
  • Priority actions — up to five fixes, ranked by severity, then by how much they affect, then by effort
  • Fix with an agent — each priority fix comes with a ready-to-run prompt. It starts a new agent in Plan mode, so nothing changes until you approve
  • 30/60/90-day plan — immediate Setup fixes first, then access and identity work, then code, endpoints and longer-term hardening
  • Health Check comparison — your Salesforce Health Check score alongside this audit, with every place they disagree
  • Filters — narrow charts and findings to Setup, Code, Guest, Privileged or Integrations, or switch the audience to show only what that reader needs
  • Searchable inventory — every Apex class, trigger, component, page and flow, showing which were analyzed and which have findings
  • Data integrity — badges that confirm the report's counts match the org

How it works

  1. Confirms your org, environment and API headroom, and sizes the org
  2. Reads your security settings and Salesforce Health Check
  3. Collects identity and access, then integrations, sharing, guest access and logging
  4. Reviews flows, scheduled jobs and your own custom code. Code from installed packages is listed but not analyzed
  5. Scores each area, ranks the priority actions and builds the 30/60/90-day plan
  6. Checks every count against the org and confirms nothing personal is exposed before building the canvas

Before you start

  • A connected Salesforce org with a System Administrator, or a user with View Setup and Configuration, View All Data and API Enabled
  • The agent asks which org to audit, whether your team works remotely or in the office, and any risks you've already accepted
  • You'll confirm before the report is published, because it holds masked security findings

A typical run takes 1–2 hours, depending on org size. Find the recipe in the Library and run it from there.

Good to know

  • The audit stops before it starts if your org is low on daily API requests
  • Accepted risks are shown as accepted, never hidden
  • Changes that can lock users out, such as locking sessions to IP, come with a recommendation to test in a sandbox first
  • Code fixes start in a sandbox and are never deployed to production without your approval
  • The score is a sorting aid, not a certification. The audit is point-in-time and read-only, not a penetration test or compliance review