The Summer ’26 Flow Updates for Admins

Summer ’26 is packed with big “Agentic Enterprise” messaging, but underneath the AI headlines, there are a few practical new features for the everyday Salesforce user who lives in Flow Builder.

‍

Officially, Salesforce highlights three main pillars: Flow Builder, Flow Orchestration, and MuleSoft for Flow (Integration) as the primary ways to “compose intelligent workflows” and seamlessly connect Salesforce to external systems.

‍

For admins and consultants, this translates into:

• More gains towards “clicks over code”: With these new Flow advancements, it’s becoming significantly easier to implement complex logic, historically requiring custom Apex or code adjustments, entirely within declarative configuration tools like Flow Builder.
  
  
• Better flow orchestration governance: Clearer, more secure end-to-end processes that allow for cleaner connectivity to external systems. Yes, this is thanks to Mulesoft.
  
  
• AI-powered flow troubleshooting: The debut of an AI Flow debugger chat to help diagnose and fix design and runtime errors quickly.

‍

In this article, I'll walk you through the most impactful Flow updates in the Summer '26 release, explaining what each feature does, why it matters, and how it can benefit Salesforce admins and consultants in their day-to-day work.

‍

Key features broken down by impact

‍

1. Logic & core enhancements

‍

• New operators in decision logic: Flow Decision elements now include 20 new date-specific operators (like Is Today, Is Tomorrow, or anniversary comparisons that ignore the year). This makes handling date criteria—like contract renewals or birthdays—much cleaner. The Admin Win: Reducing the need to build custom formula resources using MONTH() or DAY() variables just to evaluate basic calendar dates.
  ‍
  

• Environment-proof email templates: Email template references now natively survive deployments. You no longer have to manually re-link templates after pushing a flow from Sandbox to Production. The Technical Twist: Because Salesforce now stores these references by their unique API Developer Name instead of a rigid, 18-character Record ID, the connection remains perfectly intact across environments.
  
  

2. Screen flow upgrades

‍

• Hyperlinked data table lookup columns: You can now view and link directly to related records from the native Screen Flow Data Table component. By checking the new Display Options in the component sidebar, the table will display the record's actual name as a clickable hyperlink. The Clean-up: You no longer need to build custom text formula workarounds or settle for displaying a messy, unclickable 18-character Salesforce ID to your users.
  
  

• The new radio button group component: To clear up any confusion, this isn't a "select-all" multi-select tool; it is a major layout and UI upgrade. The Visual Fix: Instead of traditional, stacked radio buttons that hog massive vertical screen space, this component lets you group single-select choices into a compact, button-like responsive layout (horizontal for desktop screens, automatically wrapping vertically for mobile devices).
  
  

• Natural language screen adjustments: You can now modify screen flow layouts using natural language prompts via the Agentforce side panel. How it works: Instead of manually dragging, dropping, and configuring fields, you can simply type a command like, "Add a phone field below the email input and make it required," to speak your layout adjustments into existence.

‍

3. Agentforce integration

‍

• The "ask Agentforce" evolution – building and troubleshooting in one chat: Moving far beyond just drafting new flows from scratch, Summer ’26 expands this generative interface with a feature titled "Troubleshoot and Fix Flow Errors with Agentforce." The Big Shift: Instead of juggling separate tools, the unified Ask Agentforce panel now handles both design-time issues in your saved drafts and runtime failures from active flows. When a flow throws a fault, you use the exact same chat window to get a root-cause analysis in plain English, along with step-by-step instructions on how to fix the logic.
  
  
• Streamlined resource picker for prompts: In Prompt-Triggered Flows, the Add Prompt Instructions element gets a major usability upgrade by finally opening up the native Flow resource picker. The Fix: Previously, referencing flow variables inside prompt instructions was incredibly clunky and prone to typos because you had to type them out manually. Now, finding and inserting your resources is a standard point-and-click process.

‍

‍4. Canvas usability & performance

‍

• Collapsible fault paths: You can now visually collapse fault paths directly on the auto-layout canvas. The Practical Benefit: Enterprise-grade flows with heavy error-handling paths can quickly become a visual nightmare. Toggling these paths "closed" lets you focus entirely on your primary "happy path" workflow without endless scrolling.
  
  
• Redesigned validation panel: The validation panel has been completely overhauled to stay completely out of your way until you actually click it open. The Upgrade: Errors and warnings are no longer jammed into a tiny list; they are now organised into clean, readable cards grouped by the specific element causing the issue, making debugging a saved draft much less overwhelming.
  
  
• Custom batch sizes for scheduled flows: You can now configure custom batch sizes (ranging from 1 to 200) directly within a scheduled flow's Start element. Why this matters: Previously, scheduled flows were locked into a rigid default batch size of 200 records. Being able to lower this size is a massive win for preventing Apex governor limit timeouts when your flow is triggering heavy subflows or complex Apex plugins per record.
  
  
• Centralised value mapping for MuleSoft for flow: While MuleSoft and Flow have been able to connect via External Services, translating data values between systems was historically a headache. The Mapping Fix: If an external system sends a country code as "US" but Salesforce requires "United States," this update provides a centralised, visual mapping interface directly within the integration setup, handling the cross-system data translation without requiring a bunch of messy decision nodes or assignment elements.

‍

Takeaways 

‍

To get these features up and running in your Summer ’26 sandbox, here is exactly where to look in the Salesforce UI:

‍

• For custom batch sizes: Open a Schedule-Triggered Flow, click the Start element, and look inside the Select Object properties drawer to manually override the standard 200-record batch size.
  
  
• For clean data tables: Drop a Data Table component onto a Screen Flow canvas, open the column configuration manager for a lookup field, and check the newly added Show Record Name and Link to Record checkboxes.
  
  
• For deployable email templates: Open your existing Send Email actions on the canvas and ensure you upgrade them to Version 3.0.1 to swap out static record IDs for API Developer Names.
  
  
• For AI building and debugging: Ensure you have an Agentforce/Einstein add-on license and Data Cloud active. Once Einstein GenAI is enabled in Setup, click the Agentforce chat icon on the right vertical sidebar of the Flow canvas to prompt for screen builds or paste runtime error logs for a natural language root-cause analysis.
  
  
• For native integrations: Navigate to the Automation app and use the new Global Flow Resources setting to map inbound external values before feeding them into your Transform elements.

‍

Curious about how these new Flow capabilities could benefit your Salesforce environment? Get in touch with the Gen25 team to discuss how we can help you implement, optimise, and future-proof your automation strategy.

‍

Ryan Farnaby, Commercial Director UK

r.farnaby@gen25.com

+447368 641599

‍

Joris van Wessem, Account Executive NL

j.vanwessem@gen25.com

+31611274732