A Salesforce metadata analyzer that shows its evidence

What a metadata analyzer should tell you

Listing components is the easy part. The useful questions are about relationships: what calls this class, which Flows fire when an Opportunity closes, which fields nothing reads any more, and what a change will reach before it is deployed. MetaLens360 answers them from your org's own metadata rather than from general Salesforce knowledge, and it marks every answer with how firmly the metadata supports it.

What it reads

• Metadata inventory. A sync retrieves an org's metadata across more than 200 metadata types. The Metadata Coverage Matrix shows counts per type for every org side by side, so gaps and outliers stand out.
• Apex and triggers. Indexing parses classes and triggers into a dependency graph: callers, callees, SOQL by object, lines and complexity.
• Flows. Record-triggered and other Flows are read alongside Apex, so an execution path does not stop at the boundary between code and automation.
• Dependencies. Every component gets a dossier with its blast radius, betweenness and governor risk, and Component Presence shows which orgs contain it and where its version differs.
• Source control. Connect a GitHub, Bitbucket or Azure DevOps repository with a read-only token to compare branches with orgs and join commits to deployed metadata.

What it finds

• Execution paths. Apex Flow Analyzer traces what runs when a record changes — trigger, handlers, Flows, and queueable or Platform Event work that moves to a separate transaction — with SOQL counts and duplicate queries at each step.
• Static analysis with call-chain context. Code Analyzer follows method calls out of a loop to find SOQL or DML that is reached from it, detects re-entrancy cycles that span Apex and Flow, and lets you tune the rule catalog without a deploy.
• Flow risks. Flow health analysis ranks patterns such as DML or SOQL inside loops, missing entry criteria and after-save work that belongs in before-save.
• Dead code and orphaned automation. Components with no inbound references, no entry-point annotations and no async role are listed as candidates to delete; change-event capture that nothing consumes is listed too.
• Drift. Compare an org with a Git branch, two orgs, or two branches, and read the differences in a side-by-side diff with risk levels.

Evidence, not assertion

Measured figures — component counts, dependency edges, SOQL per transaction — come straight from the index and need no AI at all. Where MetaLens360 writes an explanation, items are tagged Observed, Inferred or Worth checking, and the copilot cites the components it relied on. If the metadata cannot answer a question, it says so instead of guessing. Unknown is reported as unknown, never as zero.

AI features run on the provider keys your workspace configures, so the model and the bill stay under your control. Everything that is measured works without a key.

How it connects to your org

MetaLens360 only reads metadata; it never deploys or changes anything in your org. You connect with an SFDX Auth URL, an access token or one-click Salesforce sign-in, run a sync, and the org is indexed for analysis. Workspaces are isolated, credentials are encrypted, and roles control who can sync, tune rules or manage connections. The first-hour checklist walks through the whole setup.

Looking for an org-level view — a health grade, technical debt and CRM Analytics lineage? See Salesforce org analysis and health check.