v6.0.0 introduces the Flow Error Explorer, a two-part feature that
helps Salesforce administrators and developers understand and resolve
runtime errors in Salesforce Flows.


When a Flow fails at runtime, Salesforce generates a fault debug link.
When Flow Builder is opened via that link, the Flow Error Explorer
automatically detects the error state, translates the raw Salesforce
fault string into plain English, and injects an Explore Error
button into the debug panel toolbar.

Clicking Explore Error opens a structured error analysis modal
showing:

• The Salesforce ExceptionCode extracted from the fault string
• A plain-English description of what the error means
• Common causes in a Salesforce Flow context
• Step-by-step remediation recommendations
• The faulted element type and label
• Any Salesforce record IDs extracted from the fault message
• A cascade failure indicator where applicable
• A one-click Copy AI Prompt action that assembles a structured
  prompt ready to paste into any AI tool

The Error Dictionary is accessible from the toolkit side button on
any Salesforce Setup page, without needing an active debug session.
It provides a browsable catalogue of known Salesforce Flow error codes
grouped by category, each with a description, causes, recommendations,
and related error codes.

Categories covered include: Validation, Governor Limits, Data
Operations, Access & Permissions, Apex & Actions, Duplicate Rules,
and Cascade Failures.



The Flow Health Check scoring model has been updated to more
accurately reflect real-world production risk.

Previously, the score deducted a flat amount per issue family
regardless of how many violations that family contained. A flow with
four DML operations inside a loop was scored identically to one with
one. This meant flows with multiple serious violations could still
score in the 70s.

The updated model deducts a base amount per family plus a per-instance
scaling charge for every violation beyond the first, capped per family
to prevent the score going negative. High-severity violations (DML
inside loops, missing fault paths on DML) compound aggressively
because each additional instance represents a real governor limit or
data safety risk. Low-severity violations (missing descriptions) remain
flat — ten missing descriptions reflects one bad habit, not ten
independent failures.

A flow designed to demonstrate bad practices — DML and queries inside
loops, missing fault paths on all data operations, outdated API version
— now correctly scores 39 (Very Poor) rather than 77 (Good).

The score remains on the same 0–100 deduct-from-100 scale, consistent
with the standard Salesforce Health Check model. The score floor is 0
and cannot go negative.