Salesforce Data 360 MCP Server Is in Developer Preview
On May 26, Salesforce announced that the Data 360 MCP Server is now in Developer Preview. The idea is direct: every piece of data in your Salesforce Data Cloud is now reachable by any AI agent that speaks Model Context Protocol. Your CRM context, unified customer profiles, real-time data streams — accessible from Claude Code, Cursor, or any MCP-compatible agent tool without writing a custom API wrapper. Data 360 MCP Server — How Data Cloud Connects to AI Agents Salesforce Data Cloud • Unified customer profiles • Audience segments • Calculated insights • Data streams • Identity resolution exposes Data 360 MCP Server Developer Preview Trust Layer protected tool calls MCP-compatible AI agents Claude Code Cursor / Codex Agentforce Studio Custom MCP clients Any agent that speaks MCP No custom API wrapper required. Data arrives through the same trust layer as other Agentforce data access. What the Data 360 MCP Server actually exposes The server gives AI agents structured access to Data Cloud data — unified customer profiles, audience segments, data streams, calculated insights, and identity resolution outputs. These are the data objects that previously required SOQL-like queries against the Data Cloud query engine or custom API development to surface. Through the MCP interface, an agent can retrieve a unified customer profile by identity, query segment membership for a specific individual, pull calculated insight values for an account, or read real-time data stream events — using the same tool-call pattern it would use to interact with any other MCP server, without learning a proprietary API. The significance is in the combination. An agent building a renewal recommendation previously had access to Salesforce CRM data — the account record, the opportunity history, the activity log. What it did not have was the Data Cloud layer: the calculated health score from product usage, the segment membership that reflects behavioural patterns, the cross-channel identity resolution that unifies how the same customer appears across touchpoints. The Data 360 MCP Server adds that layer. Why this changes how agents reason The practical difference between an agent with CRM access and an agent with CRM plus Data Cloud access is the difference between structured records and contextualised customer intelligence. An agent reviewing a renewal opportunity can currently see: account name, ACV, contract end date, last activity log, open support tickets. With the Data 360 MCP Server, that same agent can also see: the customer’s health score calculated from product usage patterns, their segment membership indicating they are in a high-churn-risk cohort, and their identity resolution confirming that two separate records in the CRM are the same individual. Because the data arrives through the MCP interface with the same trust layer protections as other Agentforce data access, the agent’s data handling governance applies uniformly — no separate security configuration needed for the Data Cloud layer. What is available in Developer Preview vs. what is coming Capability Available in Developer Preview Expected at GA Unified customer profile retrieval ✓ Query by identity, return full profile with attributes — Audience segment membership ✓ Query segment membership for a specific individual or account — Calculated insights reads ✓ Return health scores, propensity scores per record — Identity resolution queries ✓ Cross-reference unified identity across touchpoints — Data stream reads ✓ Basic event stream data — subject to Data Cloud sync latency Real-time streaming subscriptions Write-back to Data Cloud ✗ Not yet available Agents will be able to update Data Cloud records based on reasoning output Complex data stream subscriptions ✗ Not yet available Subscribe to data stream events as part of agent trigger logic Production org access ✗ Developer Edition orgs with Data Cloud only Full production org access at GA Developer Preview gives access to the core unified profile retrieval, segment queries, and calculated insight reads. The key constraint is data freshness: Data Cloud data surfaces through the MCP Server with the same refresh latency as the underlying Data Cloud sync. For most use cases this is acceptable. For agents reasoning about real-time events, it is worth understanding the lag characteristics of your specific data streams before building workflows that depend on sub-second freshness. Developer Preview access path Data 360 MCP Server — Developer Preview Access Path4 steps 1Confirm your org has Data Cloud enabledDeveloper Edition orgs with Data Cloud access qualify for Developer Preview. If you do not have a Data Cloud-enabled org, spin up a Developer Edition at developer.salesforce.com — the fastest path to experimenting. 2Enable the Data 360 MCP Server in SetupOpen Setup → search for MCP Servers. The Data 360 MCP Server appears in the available server list for Developer Preview-enrolled orgs. Enable it and configure which agents have access to which data objects.Setup → MCP Servers → Data 360 MCP Server → Enable 3Connect an MCP-compatible clientClaude Code, Cursor, Agentforce Studio, or any MCP-compatible development environment can connect once the server is enabled. The server returns a tool manifest — no custom authentication code required beyond the standard MCP handshake. 4Test with a unified profile retrievalPull a known profile by ID, confirm the MCP Server returns the same data as the Data Cloud UI, and verify that segment membership and calculated insights are included in the response. This confirms the data path is working before building agent logic on top of it. Developer Preview is the right time to experiment, not the right time to build production workflows. Map your architecture, test your data access patterns, and identify what works before GA removes the preview caveats. The orgs that move through this now will deploy faster when GA lands. Data 360MCPData CloudSalesforce DevAgentforce Share: LinkedIn Twitter / X Copy link In this article 01What it exposes 02Why it changes agent reasoning 03Preview vs. GA capabilities 04Access path — 4 steps Developer Preview status Unified profilesQuery by identity — live Segment membershipPer individual — live Calculated insightsHealth scores — live Write-backExpected at GA Production orgsDev Edition only now Quick access path 1️⃣Enable Data Cloud in org 2️⃣Setup → MCP Servers → Enable 3️⃣Connect MCP
Salesforce Summer 26 Release Features
Salesforce announced Summer ’26 on May 11 and set the general availability date for June 15. The headline: 17 major capabilities, all pointing in one direction. Agentforce is no longer a feature layer on top of the platform. In Summer ’26 it is becoming the operating layer underneath everything else. Here is what actually landed and what it means for your org. Salesforce Summer ’26 — Five Changes That Shape the Platform 🤝Multi-Agent OrchestrationAgents delegate to specialist agentsOne customer-facing contact point, multiple specialist agents working behind the scenes. Triage → delegate → coordinate — without the customer switching interfaces.GA — Summer ’26 ⚡Agentforce Self-ServiceHelp Agent in 10 clicks or fewerDeploy a Help Agent to your public website, Portal, or WhatsApp in a guided setup — designed for teams who want to trial Agentforce without a multi-week implementation.GA — Summer ’26 🛡️Security MeshUnified security fabric + risk scoringDisconnected security alerts across Service Cloud, Sales Cloud, and Experience Cloud unified into a single fabric with AI-generated risk scores. Agentforce activity included.GA — Summer ’26 🔄Flow Orchestration — FreeIncluded in Enterprise and aboveFlow Orchestration moves from a usage-limited add-on to included in Enterprise, Performance, Unlimited, and Developer editions. No usage caps, no add-on cost.Now included 📊Tableau over Model Context ProtocolAnalytics engine exposed to Agentforce agentsTableau’s analytics engine is now reachable by Agentforce agents over MCP, protected by the Agentforce Trust Layer. Agents can query revenue trends, spend analytics, and historical reports as part of their reasoning — without requiring a human to pull the report first. Closes the gap between CRM data and the BI layer.GA — Summer ’26 Multi-Agent Orchestration For most of Agentforce’s history, an agent was a single system handling a single domain. A service agent answered product questions. A sales agent qualified leads. Each worked independently and each required its own configuration. Multi-Agent Orchestration changes that architecture. Agents can now delegate tasks to specialist agents within the same org. One customer-facing contact point, multiple agents working behind the scenes: a triage agent receives the request, determines which specialist — a billing agent, a technical support agent, a returns agent — should handle it, delegates the task, and coordinates the result back to the customer without the customer ever switching interfaces. The practical implication for orgs with complex service or sales workflows is that you can build specialist agents for distinct domains and let orchestration handle the coordination, rather than trying to build one agent that knows everything. Simpler individual agents, more reliable outcomes at the orchestration level. For developers, the build model changes too. Agent teams can be tested and deployed independently. Failures in one specialist agent are contained rather than cascading through the entire interaction. Agentforce Self-Service The barrier to deploying an Agentforce Help Agent drops significantly in Summer ’26. Agentforce Self-Service is a setup path that gets a Help Agent deployed in 10 clicks or fewer — configured, grounded in your knowledge base, and ready to go on your public website, the new Portal experience, or WhatsApp. The positioning is explicit: Salesforce is targeting the orgs that have heard about Agentforce but found the implementation path too complex for their team size or technical capacity. Self-Service is designed to remove that barrier without removing the ability to customise later. For admins at smaller orgs who have been waiting for a way to trial Agentforce without a multi-week implementation project, this is the most directly actionable Summer ’26 announcement. For larger orgs, the Self-Service path is worth understanding as a rapid prototyping route before committing to a full agent build. Security Mesh Security Mesh unifies data sources across the Salesforce platform into a single security fabric and transforms disconnected access logs and alerts into intelligent risk scores. Instead of reviewing separate security events across Service Cloud, Sales Cloud, and Experience Cloud independently, Security Mesh provides a unified view with AI-generated risk assessment. The practical value for compliance-minded orgs is in audit efficiency. Security events that previously required cross-referencing multiple tools to understand their combined significance are now surfaced as correlated risk signals. Additionally, Security Mesh integrates with the Trust Layer that governs Agentforce agents, meaning agent activity is included in the unified risk picture rather than existing as a separate data source. Flow Orchestration now free Flow Orchestration moves from a usage-limited add-on to an included feature in Enterprise, Performance, Unlimited, and Developer editions without usage-based limits. This removes the licensing conversation from any multi-step process automation project. The timing is intentional. As Agentforce agents become more common in Salesforce orgs, the need to coordinate complex multi-step workflows across agents, humans, and systems increases. Flow Orchestration is the tooling that handles that coordination on the Salesforce side. Making it free removes the last friction point from adopting it broadly. For orgs that evaluated Flow Orchestration and passed because of cost, Summer ’26 is the time to revisit any approval workflows, cross-department handoff processes, or multi-system coordination tasks that are currently running on manual steps or basic Flow. Tableau over Model Context Protocol Tableau’s analytics engine is now exposed to Agentforce agents over the Model Context Protocol, protected by the Agentforce Trust Layer. An agent reasoning about a customer renewal can query Tableau for historical revenue trends. An agent managing a procurement workflow can pull spend analytics directly from the BI layer without requiring a human to run the report first. For data-heavy orgs, this is the most architecturally significant Summer ’26 addition after Multi-Agent Orchestration. It closes the gap between Salesforce CRM data — which agents have had access to — and the analytical layer that lives in Tableau but has been outside the agent’s reach. Role Most relevant Summer ’26 feature What to do now Admin Agentforce Self-Service lowers the barrier to deploying a Help Agent to your website or Portal. Flow Orchestration now free removes the licensing blocker for multi-step approval workflows. Explore Try the Self-Service setup in a sandbox. Identify one approval workflow worth rebuilding in Flow Orchestration now that cost is not a
Salesforce Time Tracking Sales
A sales manager can tell you the quota number for every rep on their team. Ask them how many hours per week those reps spend on actual selling versus admin work, meetings, and CRM updates — and the answer is usually a guess. The gap between what managers think their team is doing and what they are actually doing is where revenue goes quietly missing. True Time Tracker closes that gap, inside Salesforce, without a new tool, a new login, or a new workflow. Actual selling: 27% Typical B2B Sales Rep Time Split Actual selling activity27% Admin work and CRM updates28% Internal meetings19% Non-selling email and comms17% Other (travel, training, etc.)9% Source: Salesforce “State of Sales” research benchmarks. Your team’s split may vary — that is exactly the point. Why this gap exists at the 20 to 50-person stage At 10 people, a sales manager knows what every rep is doing because they are next to them. At 20 to 50 people, that changes. Reps are distributed, partially remote, or simply moving fast enough that day-to-day time patterns are invisible to leadership. Most sales tools track outcomes — deal value, close rate, pipeline stage. None of them track inputs in a way that is actionable. Pipeline reports tell you what happened. Time data tells you why it happened and what is likely to happen next. Without time visibility, the only lever a manager has when results are underperforming is to ask the rep what they think the problem is. That is a useful conversation but not a reliable diagnostic. Three decisions that become better with actual time data Account coverage Time allocation visibility lets managers see the full picture: a rep spending 14 hours a week on three legacy accounts — accounts that are renewing at stable rates and require minimal active management — while high-potential accounts get two hours each. The result shows up in pipeline three months later, not in this week’s activity log. With time data in Salesforce, the conversation changes from “why are these deals not progressing” to “let us look at where the time is going and redistribute it deliberately.” That is a more productive conversation with a more actionable outcome. Coaching conversations Most coaching conversations in sales are about results: close rate, pipeline coverage, deal velocity. These are lagging indicators. They tell you what already happened. Time patterns are leading indicators. A rep spending 60 percent of their selling time on proposals and zero time on prospecting will have an empty pipeline in six weeks. A rep who has not had a discovery call with a new prospect in 14 days is building the same problem. Time data surfaces these patterns before the pipeline report does. Headcount decisions Before hiring the next sales rep, most companies look at pipeline coverage and close rates. The more direct question is: how are current reps actually spending their time, and where are they constrained? If reps are spending three hours a day on admin that could be automated or systematised, the capacity problem is not a headcount problem. If they are spending all available hours on active selling and the pipeline still cannot grow, it is. Time data makes the distinction visible before a hiring decision is made. Decision Without time visibility With True Time Tracker Account coverage Manager asks rep why deals are not progressing. The real issue — hours disproportionately allocated to low-ARR accounts — is invisible. Manager sees actual time per account vs ARR per account. Reallocation conversation is data-driven: “You spent 14 hours on these three accounts. Here is what the time looks like vs revenue potential.” Rep coaching Coaching is based on close rate, pipeline coverage, deal velocity — lagging indicators. Manager reacts to what already happened. Coaching is based on time patterns — leading indicators. Rep spending 60% of time on proposals and 0% on prospecting will have an empty pipeline in 6 weeks. Manager can see and address this now. Capacity planning Headcount decision based on pipeline coverage and close rates. Team looks busy. Manager hires another rep. Productivity problem continues. Time data shows reps spending 3 hours/day on admin that could be systematised. Bottleneck is process, not people. Automation before hiring saves the cost of a rep. Pipeline forecasting Manager estimates based on rep self-reporting. Forecast accuracy is moderate at best and degrades as the quarter progresses. Time patterns on high-value accounts are a leading indicator of deal velocity. Time data improves forecast input quality before the pipeline report catches up. How True Time Tracker works inside Salesforce True Time Tracker logs time natively inside Salesforce, against the records that time relates to: Opportunities, Accounts, Activities, or custom objects. Reps log time in the same interface they use to update deals. Managers see time data in dashboards alongside pipeline data, without switching tools. The most common objection to time tracking is rep resistance — a perception that it is surveillance rather than a management tool. True Time Tracker addresses this by making the data visible to the rep as well as the manager. Reps can see their own time patterns, which is often the most effective way to surface inefficiencies that they were not aware of. Three questions True Time Tracker answers that your pipeline report cannot Pipeline reports track results. Time data tracks what produces them. 1Are my reps spending time on the right accounts?Time per account vs ARR per account reveals coverage misalignment immediately. High-potential accounts receiving low time allocation show up clearly — weeks before the missed deal shows up in pipeline.Pipeline report answer: “Here are the deals and their stages.” — Not useful for spotting coverage problems. 2What is actually taking up my reps’ selling time?If a rep’s capacity is constrained, the question is whether it is constrained by selling activity or by admin and meetings. Time data gives you that breakdown. The intervention is different depending on the answer.Pipeline report answer: “The rep has 12 open opportunities.” — Does not tell you why
Salesforce Summer 26 Admin Checklist
The Summer ’26 production upgrade is landing on June 5 and June 12 for most Salesforce orgs. Sandboxes have been on preview since May 8, which means there is no excuse for surprises on production upgrade weekend. Here is the pre-upgrade checklist every admin needs before their org flips. Summer ’26 Production Upgrade Weekends 🚨 June 5 — TomorrowMain production waveMost orgs on NA, EU, and AP instances. If your org is on this wave, the checklist below needs to be complete today — not this weekend. 📅 June 12 — 8 daysFinal production waveRemaining instances. You have until June 11 EOD to complete all mandatory items. Use the time — do not save this checklist for June 10. The mandatory items — these break things if skipped 1. SAML migration New SAML defaults are enforced in Summer ’26. If your org uses SSO via SAML — whether Salesforce is the identity provider, the service provider, or both — you need to verify your Auth Provider configuration and test the full login flow in your Summer ’26 sandbox before production upgrade. A SAML configuration that worked in Spring ’26 may fail after the upgrade if the new defaults require updated settings. Additionally, Triple DES signing for SAML SSO stops working entirely in Summer ’26, as announced in Spring ’26. Any SAML configuration using Triple DES as the signing algorithm breaks on upgrade regardless of whether you took any other action. 2. Apex sharing and security defaults New Apex security behaviors are enforced. Custom Apex code that relies on specific sharing model assumptions from earlier releases needs review. If your org has custom Apex classes managing record-level access or security, test them against the Summer ’26 sandbox before your production upgrade date. 3. Standard Omni-Channel retirement Standard Omni-Channel was retired June 1. If your production org has not migrated to Enhanced Omni-Channel Routing, this is now urgent — not scheduled. Completing the migration before your production upgrade date is the priority. See the dedicated Omni-Channel migration article for the correct four-step sequence — channel migrations must happen before the routing switch is enabled. 4. Legacy PDF generation retirement A legacy PDF rendering behavior is retired in Summer ’26. Any process that generates PDFs from Salesforce — Quote PDFs, Visualforce-generated reports, any custom PDF output — should be tested in the Summer ’26 sandbox to confirm the output is correct before production upgrade. PDF rendering changes are difficult to detect until something generates incorrectly in front of a customer. Summer ’26 Pre-Production Upgrade ChecklistBefore June 5 or June 12 Mandatory — breaks things if skipped SAML auth provider configuration verifiedMandatoryNew SAML defaults enforced. Test full SSO login flow in Summer ’26 sandbox. Triple DES signing algorithm stops working entirely — update to SHA-256 if still using Triple DES.↳ Breaks: SSO login fails for all SAML-authenticated users after upgrade Custom Apex sharing code reviewedMandatoryNew Apex security defaults enforced. Custom Apex classes managing record-level access or sharing rules need testing in sandbox against the new defaults. Run Apex tests in the Summer ’26 sandbox before production upgrade.↳ Breaks: Record access violations or unexpected sharing behaviour in Apex-governed objects Enhanced Omni-Channel migration completeMandatory · UrgentStandard Omni-Channel retired June 1. Migrate service channels (Live Agent, SMS, Messenger) first, then enable Enhanced routing. Test agent login and work item routing in sandbox end-to-end.↳ Breaks: Agents cannot log in to Omni-Channel, work items stop routing entirely PDF generation processes tested in sandboxMandatoryLegacy PDF rendering behaviour retired. Any process generating PDFs from Salesforce — Quote PDFs, Visualforce reports, custom PDF output — must be tested in Summer ’26 sandbox.↳ Breaks: Incorrectly formatted or failed PDF generation for customer-facing documents Recommended — should be done; will not immediately break Agentforce agent configurations reviewedRecommendedSummer ’26 changes agent lifecycle defaults. If your org has agents in production, review the Agentforce Summer ’26 release notes and test agent initialisation, session handling, and error surfacing in sandbox. Evaluate Flow Orchestration for existing processesOptional upsideFlow Orchestration is now free in Enterprise, Performance, Unlimited, and Developer editions. Review multi-step approval processes or cross-object coordination workflows that could benefit from rebuilding in Orchestration. Exact production upgrade date confirmedRecommendedSetup → Company Information → Instance. Then trust.salesforce.com to find your exact upgrade weekend. Some instances upgraded May 15 — if yours was in that wave, your production org is already on Summer ’26. The items you should do but that will not break immediately 5. Flow Orchestration is now free Flow Orchestration is included in Enterprise, Performance, Unlimited, and Developer editions without usage limits in Summer ’26. If your org has been avoiding Flow Orchestration due to licensing cost, that barrier is gone. Evaluate whether any current multi-step approval processes or cross-object coordination workflows would benefit from rebuilding in Orchestration. 6. Review Agentforce configurations Summer ’26 changes agent lifecycle defaults for orgs running Agentforce in production. If your org has agents in production — even in early access or limited deployment — review the Summer ’26 Agentforce release notes and test agent behaviour in sandbox before production upgrade. 7. Confirm your exact upgrade date Do not assume June 5 or June 12. Your specific upgrade date depends on your Salesforce instance. Check: Setup → Company Information → Instance field. Then confirm your upgrade date at trust.salesforce.com. Some instances upgraded May 15. If your org is on one of those instances and you are reading this on June 4, your production org may already be on Summer ’26. Summer ’26 is the release where “I’ll check the sandbox later” becomes a problem. The June 5 wave is tomorrow. If your sandbox has been on preview since May 8 and you have not looked at it yet, today is the day. Salesforce Admin Summer ’26 Salesforce Release SAML Upgrade Checklist Share: LinkedIn Twitter / X Copy link In this article —Mandatory items 01SAML migration 02Apex security defaults 03Omni-Channel migration 04PDF generation —Recommended items 05Flow Orchestration free 06Agentforce review 07Confirm your date Upgrade dates May 15First wave — already liveCheck if
Standard Omni-Channel Retirement Salesforce
Today is June 1. Standard Omni-Channel in Salesforce is officially retired. If your org has not migrated to Enhanced Omni-Channel Routing and your production instance hits the Summer ’26 upgrade before you are ready, your agents will not be able to log in to Omni-Channel and work items will stop routing entirely. This is not a beta warning. This is now. ⚠What breaks if your org upgrades without completing the migration Agent loginAgents see an error when attempting to log in to Omni-Channel. They cannot set themselves as Available and cannot receive routed work items until the migration is completed. Work routingIncoming cases, chats, and calls stop routing entirely to any queue using Standard Omni-Channel logic. Work items queue without assignment until an admin completes the migration. Specific channelsLive Agent (Chat), standard SMS, and Facebook Messenger queues are the highest-risk channels. Each requires individual migration before Enhanced routing is enabled — enabling Enhanced routing first on these breaks them even if the main switch was working. ResolutionThe fix requires completing migration steps in production under time pressure, potentially during a service outage. The correct time to do this is now, in a sandbox, not after the upgrade breaks routing. Step 1: Check your current Omni-Channel status Open Setup and search for Omni-Channel Settings. If your org shows Standard Omni-Channel Routing as active, you have not completed the migration. If Enhanced Omni-Channel Routing is enabled, check whether all service channels have been migrated — it is possible to have Enhanced routing active but individual channels still on the legacy configuration. The specific channels that require individual migration steps before you enable Enhanced routing are: Live Agent (Chat), standard SMS channels, and Facebook Messenger. Enabling Enhanced Omni-Channel routing before migrating these channels will break routing for those specific queues. Step 2: Check your production upgrade date Not all production orgs upgrade on the same weekend. The Summer ’26 production upgrade runs across three windows: some instances upgraded May 15, the main production wave is June 5, and the final wave is June 12. Your upgrade date determines how much time you have. Find your instance: Setup → Company Information → Instance field. Then check your specific upgrade date at status.salesforce.com. If your instance upgrades June 5, you have four days from today. If it upgrades June 12, you have eleven. How to find your production upgrade date 1Open Salesforce Setup → search for Company InformationSetup → Company Information → Instance 2Note the instance name (e.g. NA87, EU15, AP5) 3Go to trust.salesforce.com → select your instance → find the Summer ’26 maintenance windowtrust.salesforce.com/status/maintenance Summer ’26 Production Upgrade Waves May 15First wave — selected instances already upgraded. If your org is on this wave, you are already on Summer ’26.Already live June 5Main production wave — majority of orgs. This is Thursday. The window to prepare is today.Tomorrow June 12Final wave — remaining instances. You have until June 11 EOD if your instance is in this wave.7 days left Step 3: Migrate in the correct sequence Enhanced Omni-Channel Migration — Correct SequenceDo in this order 1Migrate individual service channels firstLive Agent (Chat), standard SMS channels, and Facebook Messenger must be individually migrated before you enable Enhanced routing. Enabling Enhanced routing before these migrations breaks them. This is the step most orgs get wrong.Do this before anything else 2Enable Enhanced Omni-Channel RoutingAfter channel migrations in Step 1 are complete, open Omni-Channel Settings in Setup and enable Enhanced Omni-Channel Routing. This is the main routing switch — only flip it after Step 1 is confirmed complete.Only after Step 1 3Verify agent capacity model settingsEnhanced routing uses capacity-based logic that may differ from your Standard configuration. Review queue settings and capacity model assignments in your Summer ’26 sandbox. Confirm routing rules behave as expected with simulated work items.In sandbox first 4Test full agent login and work item routing end-to-endHave a test user log into Omni-Channel in the Summer ’26 sandbox, set status to Available, and accept a routed work item through the full flow. Confirm the item closes and activity is logged correctly before applying these settings to production.Verify before production Step 4: Check for exemptions Two specific scenarios may affect your migration timeline. If your org has a legacy Chat implementation that qualified for a legacy-chat exemption, you may have additional transition time — check your org’s Salesforce communication history for any exemption notification. If your org has not yet agreed to Hyperforce AWS terms, certain Enhanced Omni-Channel features may be unavailable until those terms are accepted. If your production org upgrades before this migration is complete: agents will see an error when attempting to log into Omni-Channel, and work items will not route. The fix requires completing the migration steps under time pressure in a production environment. Completing them now, in a sandbox, is the better version of this conversation. Salesforce Admin Omni-Channel Summer ’26 Service Cloud Share: LinkedIn Twitter / X Copy link In this article 01Check your Omni-Channel status 02Find your upgrade date 03Migrate in the correct sequence 04Check for exemptions ⚠ What breaks if you skip this Agents can’t log in to Omni-Channel All work items stop routing Chat, SMS, Messenger break individually Fix in production = outage pressure Migration sequence 1Migrate channels first 2Enable Enhanced routing 3Verify capacity model 4Test end-to-end in sandbox About the Author ST Sergey Trusov CEO & Salesforce Architect at TrueSolv
Salesforce Q1 FY27 Earnings Preview
Salesforce reports Q1 FY27 earnings on June 3. Last quarter: $11.2B in revenue, 29,000 Agentforce deals closed, $800M in Agentforce ARR — and guidance for continued Agentforce-led growth. The question everyone will be watching is not whether the revenue number grew. It is whether Agentforce ARR is accelerating and how many of those 29,000 deals turned into real deployments. Salesforce FY26 Full-Year Results — The Baseline for June 3 Q1 FY27 earnings reported June 3, 2026 after market close $41.5B+10% YoYFY26 full-year revenue — highest annual total in company history $800M+169% YoYAgentforce Annual Recurring Revenue at end of FY26 29,000+50% QoQAgentforce deals closed since launch — commercial and public sector $72BRPORemaining Performance Obligations — contracted future revenue Source: Salesforce FY26 Q4 Earnings, February 25, 2026. Q1 FY27 results on June 3 are the first post-FY26 signal on whether Agentforce momentum is accelerating or plateauing. Watch 1: Agentforce ARR trajectory FY26 closed with $800M in Agentforce ARR after 169 percent year-over-year growth. Q1 FY27 is the first full quarter with three key products in market simultaneously: Agentforce Sales went GA on March 16, Agentforce Operations went GA on April 29, and Agentforce Contact Center is live in Enterprise and Unlimited editions. The ARR number on June 3 is the first clean signal of whether enterprise adoption is compounding from the FY26 base or plateauing as initial deal signings convert into measured deployments. A significant step-up from $800M suggests acceleration. Flat or modest growth suggests the 29,000 deal count is still primarily pilots and signed agreements rather than active production deployments. Additionally, watch the combined Agentforce and Data Cloud ARR figure. In FY26, that number exceeded $2.9B. The Data Cloud layer is the data substrate that makes Agentforce agents reliably useful — its trajectory tells you something about the depth of enterprise adoption beyond surface-level AI feature adoption. Watch 2: Deployment signals versus deal count Twenty-nine thousand Agentforce deals signed is a pipeline number. The more interesting metric is what proportion of those deals moved from signed to live in production. Salesforce provided proxy signals for this in FY26 — token consumption (nearly 20 trillion tokens processed) and agentic work units (2.4 billion delivered) — as evidence of real operational output rather than just signed contracts. Q1 FY27 will either extend those proxy metrics significantly or provide a more cautious signal about deployment pace. Token consumption accelerating quarter-over-quarter is the clearest indicator that the deal count reflects real production usage, not pipeline optimism. Three Questions to Watch on June 3Earnings preview 1 Is Agentforce ARR accelerating from the $800M FY26 base? Q1 FY27 is the first full quarter with Agentforce Sales (GA March 16), Operations (GA April 29), and Contact Center all in market. A significant step-up signals compounding enterprise adoption. Flat growth signals deals are still converting slowly from signed to deployed. Bullish signal: ARR significantly above $800M run rate 2 Are the proxy deployment metrics (tokens, agentic work units) accelerating? FY26 reported nearly 20 trillion tokens and 2.4 billion agentic work units — operational evidence of real production usage. If these numbers step up materially in Q1 FY27, it confirms a meaningful proportion of the 29,000 deals are live in production. Bullish signal: token consumption and agentic work units significantly higher 3 Any signal on Agentforce traction below the enterprise segment? FY26 Agentforce growth was primarily enterprise-led. The Spring ’26 release — AgentExchange consolidation, Salesforce Setup for SaaS, managed package templates — signals intent to accelerate mid-market and SMB adoption. Commentary on sub-enterprise traction would be significant. Watch for: SMB and mid-market Agentforce references in prepared remarks Watch 3: SMB and mid-market traction FY26 Agentforce growth was predominantly enterprise-led. Large deals with named enterprise customers drove the majority of the ARR. The Spring ’26 release — including the Salesforce Setup for SaaS initiative, the AgentExchange marketplace consolidation, and the acceleration of managed package templates for specific verticals — signals intent to bring Agentforce adoption into the mid-market and SMB segments. Salesforce’s $41.5B revenue base was built primarily on SMB and mid-market customers. The long-term Agentforce story depends on whether the platform can deliver agent value at that tier, not just at the enterprise level where implementation complexity is more manageable. One more thing: the Earnings Show format Salesforce moved its earnings calls to a more informal ‘Earnings Show’ format that often includes customer CEO guests and a conversational structure alongside the traditional financial presentation. It is worth watching in full rather than reading the transcript — the customer case studies and Benioff’s commentary on platform direction often contain more signal about where the product is going than the prepared remarks alone. 📺 About the Salesforce Earnings Show — June 3 🕔Time: After market close on June 3, 2026. Typically begins 1 hour after close with the press release, followed by the live show. 🎙️Format: Conversational structure alongside traditional financial presentation. Often includes customer CEO guests discussing real deployment outcomes. 📊Beyond the numbers: Benioff’s commentary on Agentforce deployment depth, customer case studies, and any commentary on the SMB and mid-market motion. 🔗Where: investor.salesforce.com — live stream and replay. TrueSolv will be covering the call live on LinkedIn. The revenue number on June 3 will tell you how Salesforce is doing. The Agentforce ARR trajectory and the deployment signals will tell you whether the platform bet is compounding. Those are different questions and the second one matters more for anyone who depends on Salesforce as infrastructure. Salesforce Earnings Q1 FY27 Agentforce Salesforce News CRM Share: LinkedIn Twitter / X Copy link In this article 01Watch 1: Agentforce ARR trajectory 02Watch 2: Deployment vs. deal count 03Watch 3: SMB & mid-market 04The Earnings Show format FY26 baseline — key numbers $41.5BFY26 full-year revenue $800MAgentforce ARR (end of FY26) 29KAgentforce deals closed $72BRemaining Performance Obligations $2.9BAgentforce + Data Cloud combined ARR June 3 — what to watch 📈Agentforce ARR step-up from $800M ⚙️Token consumption acceleration 🏢Sub-enterprise traction signals 📺Earnings Show — watch in full About the Author DS Daria Savelieva Salesforce Consultant &
Salesforce Field Level Security Summer 26
Any Salesforce admin who has ever done a field-level security audit knows the process: open a profile, navigate to object settings, find the field, note the access, repeat for every other profile, then repeat for every permission set. It is tedious, error-prone, and nobody does it as often as they should. Summer ’26 adds a Field Access tab directly to Object Manager that shows the full picture in one place. What the Field Access tab shows At the bottom of each object in Object Manager, a new Field Access tab lists every field on the object alongside a consolidated view of exactly how access is granted — across all profiles and permission sets — in a single interface. Previously, getting this view required navigating each profile individually, cross-referencing permission sets separately, and building a mental or spreadsheet-based picture of who can see and edit which field. For a mid-sized org with 20 profiles and 40 permission sets, that process takes hours and is almost guaranteed to miss something. The Field Access tab shows it in one view. One object, every field, all access configurations visible simultaneously. ⚡ Salesforce Setup → Object Manager → Opportunity → Field Access Opportunity Standard Object Details Fields & Relationships Page Layouts Validation Rules Field Access New All profiles and permission sets — one view Field Label Profile / Permission Set Access Granted Via Amount ✓ Read / Edit Sales Rep, Sales Manager Profile: Sales Rep · Profile: Sales Manager Annual Contract Value ✓ Read / Edit Finance, Admin📖 Read only Sales Rep Perm Set: Finance View · Profile: Admin Internal Deal Notes ✓ Read / Edit Sales Manager, Admin✗ No access Sales Rep, Support Perm Set: Manager Access · Profile: Admin Stripe Subscription ID 📖 Read only Finance, Admin✗ No access Sales Rep, Support, CS Perm Set: Finance View · Profile: Admin Read-only in Summer ’26. Edit permissions via Profile or Permission Set settings. Why this matters more than it sounds Field-level security is one of the most common gaps in Salesforce org audits. Orgs grow, permission sets multiply, profiles get copied from other profiles, and nobody has a clear picture of who can read or edit which field. Compliance audits, security reviews, and new admin onboarding all require this visibility — and getting it has always required more effort than it should. The Field Access tab does not change permissions. It makes the existing permissions visible and auditable at a glance. That distinction matters for compliance contexts specifically: the audit requirement is often to demonstrate that someone reviewed field access, not that they changed it. Summer ’26 makes that review faster, more reliable, and easier to document. Three situations where this saves significant time ✗ Before Summer ’26 1Open Profile 1 → Object Settings → navigate to field → note access level~3 min per profile 2Repeat for all 20 profiles. Build a spreadsheet.~60 minutes 3Open each permission set and check the same field. Add to spreadsheet.~30 min for 40 perm sets 4Cross-reference manually. Identify gaps. Probably miss one.~20 minutes 2+ hours. Error-prone. Often skipped. ✓ With Summer ’26 Field Access tab 1Open Object Manager → select object → click Field Access tab~30 seconds 2All fields listed. All profiles and permission sets in a single view.Immediate 3Search or scroll to the specific field. Access visible at a glance.~2 minutes 4Screenshot or export for the audit documentation. Done.~5 minutes total Under 10 minutes. Reliable. Auditable. Where the Field Access Tab Saves Significant Time 🔒Before a new team accesses sensitive account dataWhen a new department or external partner is being onboarded to Salesforce, reviewing field-level security on Account, Contact, or Opportunity objects is a required step. The Field Access tab makes this a 10-minute review instead of an afternoon, and makes the outcome documentable.Time saving: ~2 hours → ~10 minutes per object reviewed 📋GDPR and HIPAA-adjacent field visibility auditsDemonstrating controlled access to specific fields — email addresses, phone numbers, health-related custom fields — requires showing who has access and how it is granted. The Field Access tab produces that view without a manual cross-referencing exercise, making it audit-ready by design.Compliance requirement: access visibility is demonstrable in one screenshot 🔍Debugging a field not visible for a specific userA rep reports a field is missing from their record page. Previously: check profile → check permission sets → check page layout. With Field Access tab: search for the field, see all access configurations simultaneously, identify the gap in under 2 minutes.Debugging time: ~20 minutes → ~2 minutes What it does not do yet The Field Access tab is read-only in the initial Summer ’26 implementation. You can view field access across all profiles and permission sets, but you cannot edit permissions from this interface. Changes still require navigating to the profile or permission set and making edits there. Worth watching: The ability to edit permissions from the same view — which would make it genuinely powerful — is the likely Winter ’27 addition. The visibility improvement is real and significant now. Editing from the same surface is the logical next step. Field-level security has always been one of the hardest things to audit in a Salesforce org. Summer ’26 does not solve the permissions complexity — it makes it visible. That is the starting point for everything else. Salesforce Admin Summer ’26 Field Level Security Object Manager Salesforce Release Share: LinkedIn Twitter / X Copy link In this article 01What the Field Access tab shows 02Why this matters 03Three time-saving scenarios 04What it doesn’t do yet Field Access tab — quick facts 🆕Where: Object Manager → any object → Field Access tab (last tab) 📊Shows: All profiles and permission sets in one consolidated view 🚫Doesn’t do: Editing — read-only in Summer ’26 📅Available: Summer ’26 GA — sandboxes from ~May 9 Use this tab when… 🔒Onboarding a new team to sensitive data 📋Running a GDPR / HIPAA audit 🔍Debugging a missing field About the Author DK Dilyara Kolesnikova Salesforce Developer & Technical Writer at TrueSolv
How to Use Agentforce to Automate SaaS Renewal Protection
Most SaaS companies do not lose renewals because the product failed. They lose them because nobody noticed the signals in time — usage dropping, a key contact going quiet, a support ticket that never fully resolved. By the time the CS team follows up, the customer has already decided. Agentforce inside Salesforce changes this from a reactive problem to a proactive workflow. Proactive CS motions outperform reactive ones at every stage The signal arrives weeks before the decision. The question is whether your system is watching for it. 67%Higher renewal rate when CS reaches out proactively vs. waiting for the customer to initiate 14 daysAverage time between detectable usage decline and churn — the intervention window most teams miss 80+Accounts per CS rep at a 30-person SaaS company — volume that makes manual monitoring impossible SaaS industry renewal benchmarks. Specific percentages vary by product category, deal size, and CS team structure. Part 1: Why SaaS renewal churn is a data problem, not a people problem The CS team at a 30-person SaaS company is usually one or two people managing 80 or more accounts. They are good at their jobs. They are not able to proactively review every account monthly while simultaneously handling onboarding calls, support escalations, and renewal negotiations. The signals that predict churn are not hidden. Usage declining by 40 percent over two weeks is visible in your product analytics. A key contact going silent for 45 days is visible in your activity log. A support ticket that stayed open for 12 days before resolution is visible in your case management. The problem is that nobody is watching all of these signals simultaneously across 80 accounts. Agentforce agents can. They run continuously against your Salesforce data, evaluate conditions against defined thresholds, and take action when those thresholds are crossed — without requiring a CS rep to remember to check. The data advantage compounds over time. As agents surface patterns — which usage drops correlate with churn, which onboarding milestones predict expansion — your renewal playbook becomes more precise with each cycle. Part 2: What an Agentforce renewal agent actually does An Agentforce renewal agent is an automated system that monitors account health signals, evaluates them against configured thresholds, and takes a defined action when a threshold is crossed — without human initiation. Agentforce Renewal Workflow — How a churn signal becomes a CS action Data Signals • Usage −40% / 14d • No contact 45+ days • Open support tickets • Contract 60d away Monitors Agentforce Renewal Agent Evaluates all signals simultaneously • 24/7 Triggers Autonomous Actions Priority CS task assigned to account owner Account summary usage + contacts + tickets Churn risk flag renewal dashboard alert Draft renewal email queued for CS review No human initiation required. Agent runs continuously against Salesforce data. The distinction from a standard Salesforce Flow is the reasoning layer. A Flow fires when a condition is met. An Agentforce renewal agent evaluates the condition in context — weighing multiple signals simultaneously, generating a natural language summary of what it found, and drafting context-aware communication that a Flow cannot produce. Part 3: Three renewal agent workflows worth building first Workflow 1 60-day renewal outreach agent Trigger:Contract end date is 60 days away. What it does:Reviews account health signals — product usage trend over the last 30 days, open support issues, last contact date, any logged expansion signals. Drafts a personalised renewal email with specific context from the account record. Queues the draft for CS review with a task to approve or revise within 48 hours. Why it works:A calendar reminder tells the rep to follow up. This agent tells the rep what the account looks like right now, drafts the message, and makes the rep’s job to review and send — not to research and write. The rep’s 20 minutes of prep becomes 5 minutes of review. Expected outcome: Higher quality renewal outreach, earlier in the cycle, with no additional CS capacity required. Workflow 2 Churn risk alert agent Trigger:Product usage drops more than 40 percent over a 14-day window compared to the prior 14 days. What it does:Flags the account with a Churn Risk tag in Salesforce. Creates a priority task for the account owner. Generates an account summary: last contact date, open support issues, usage trend, contract value and end date, any recent expansion or contraction signals. Why 14 days:The most actionable churn signals happen 30 to 60 days before renewal. A 14-day usage decline detected at 60 days out gives the CS team time to intervene before the customer has made a decision. Detected at 5 days out, the same signal is too late. Expected outcome: CS team prioritises the highest-risk accounts with full context before the intervention window closes. Workflow 3 Post-onboarding health agent Trigger:New subscription start date is 30 days ago. What it does:Checks whether the account has hit three key activation milestones. If all three are met, logs the health check and takes no further action. If one or more are not met, creates a targeted outreach task with a guided onboarding checklist attached and flags the account for a CS review. Why it matters:Customers who do not activate in the first 30 days are significantly more likely to churn at their first renewal. A health check at 30 days identifies at-risk customers at the moment where intervention — a 20-minute call, a targeted email with specific steps — is most likely to work. Expected outcome: Improved 30-day activation rates, lower first-renewal churn, CS effort focused on accounts that need it. Agent workflow Trigger What the agent does Expected outcome 📅 60-day renewal outreach Contract end date is 60 days away Reviews account health, drafts personalised renewal email, queues for CS review with 48-hour approval task Higher quality, earlier renewal outreach. Rep prep: 20 min → 5 min review. ⚠️ Churn risk alert Product usage drops 40%+ over 14-day window Flags account, creates priority task, generates account summary with full context CS prioritises at-risk accounts with
Eight Summer ’26 Features Salesforce Developers Should Actually Care About
Summer ’26 release notes landed April 22. Sandboxes upgrade around May 9. If you have not looked at the notes yet, here are the eight features most worth your time — ranked by how much they will actually change your day-to-day development workflow, not by how impressive they sound in the release announcement. Feature Impact Action before May 9 sandbox upgrade LWC Single Component Preview High · GA Test with your most complex custom component. Verify preview correctly reflects nested child component states. No production risk — pure upside for iterative development speed. LWC State Management High · GA Identify pages with multiple LWC components sharing state via prop drilling or message channels. Build a proof-of-concept state store in sandbox before committing to a refactor plan. Collapsible Fault Paths in Flow Medium · GA No action needed before upgrade. Open your largest production flow in the Summer ’26 sandbox to confirm the canvas renders correctly and collapse fault paths for readability. Global Flow Resources Medium · Preview Enable in a preview sandbox. Identify the variables declared repeatedly across your flows. Evaluate the scope of what a GA version could simplify — but do not plan production changes until GA. AI Content Summarizer Component Medium · GA Drop the component onto an Opportunity or Account page in sandbox. Verify Agentforce has data access to the fields that would make the summary useful. Flag gaps for configuration before production. External Client Apps Enforcement High · Act now Open App Manager now. Filter by Connected App type. Identify every Connected App your org created and manages. Confirm migration status to External Client Apps before Summer ’26 tightens enforcement. Web Console for Apex/SOQL Medium · GA Familiarise yourself with the interface in sandbox. Evaluate whether it replaces Developer Console for ad-hoc work or supplements VS Code. Low risk — no configuration required. Agentforce in CRM Surfaces High · Test now Load your production-critical customised pages in the Summer ’26 preview sandbox. Check for layout conflicts between custom Lightning components and new Agentforce-aware UI elements before the upgrade lands in production. 1. LWC Single Component Preview This is the feature that has appeared on Salesforce developer wishlist surveys for multiple consecutive years and finally reached GA in Summer ’26. The ability to preview a single Lightning Web Component in the browser or in VS Code without triggering a full page reload is not a small improvement — it removes one of the most consistent time sinks in iterative LWC development. The practical impact depends on how much LWC work you do. For a developer spending four hours on a component UI, the difference between a 90-second page reload and an instant preview compounds significantly across a day. For teams building complex component libraries, the impact is considerable. GA means it is stable, supported, and safe to rely on in production development workflows. Test it in your Summer ’26 sandbox first to confirm it behaves correctly with your component architecture. 2. LWC State Management State Management for LWC reaches full general availability in Summer ’26, providing centralised state across component trees — the standard pattern for managing shared state in modern frontend development that LWC has been missing. The practical scenario: an Opportunity detail page has a line items component, a totals component, and a discount component. All three need to respond to the same underlying deal data. Without State Management, this requires lifting state through the parent component via properties, creating a coordination pattern that grows increasingly unwieldy as component trees deepen. With State Management, each component subscribes to a shared store directly. Before — prop drilling across components // Parent must hold all shared state // and pass down as props to each child @track totalMrr = 0; @track discount = 0; @track finalPrice = 0; // Wired to LineItems via: <c-line-items total-mrr={totalMrr} discount={discount}> </c-line-items> // And to Totals separately… // Grows unwieldy fast. After — centralised state store // Shared store — declared once import { createStore } from ‘@salesforce/state’; export const oppStore = createStore({ totalMrr: 0, discount: 0, finalPrice: 0 }); // Each component subscribes directly // No parent coordination needed // LineItems, Totals, Discounts // all update from the same store 3. Flow UI: Collapsible Fault Paths and Readable Data Tables Collapsible fault paths extend the canvas cleanup work started in Spring ’26 with collapsible Decisions and Loops. Large production flows become difficult to maintain when fault paths branch from every element that can fail and the canvas is an unnavigable tangle of error-handling logic. For developers maintaining flows built by previous admins, collapsible fault paths are a significant cognitive load reduction. The execution path becomes readable when error handling is collapsed. When debugging, expand the specific fault path you are investigating rather than navigating around all of them simultaneously. The data table display improvements in Flow Builder are in the same category — cosmetic in that they do not change flow behaviour, but meaningful for any developer who has tried to read a data table in the current canvas and given up. 4. Global Flow Resources Still in preview orgs, but worth testing now: Global Flow Resources enable variables and components to be shared across flows platform-wide rather than redeclared in every flow that needs them. The impact depends on your org’s automation architecture. Orgs with dozens of flows that each declare the same custom picklist variable or the same error-handling configuration are the immediate beneficiaries. If this ships as expected in Winter ’27 based on the current preview, it significantly reduces duplicated logic across complex automation layers. Because it is in preview, it belongs in your summer sandbox testing but not in your production planning until GA is confirmed. 5. AI Content Summarizer Component A new AI Content Summarizer component can be dropped onto any Lightning record page from App Builder without writing Apex. It surfaces an Agentforce-generated summary of the record’s key information — account history, deal context, case background — directly on the page. The developer
Agent Script: The Layer Between AI Flexibility and Enterprise Control
The biggest complaint about AI agents in production has always been the same: they work great in demos and go off-script in real deployments. Salesforce heard that complaint and built Agent Script — a scripting language that lets you define exactly how an agent behaves, step by step, without giving up the flexibility that makes AI useful. Here is what it is, how it works, and why it changes the calculus on Agentforce adoption. Agent Behaviour: Without vs. With Agent Script WITHOUT Agent Script — probabilistic path WITH Agent Script — deterministic rails User request Path A Path B Path C Outcome depends on reasoning and conversation flow ⚠ Risk: agent may skip required verification if the user’s phrasing is persuasive or impatient User request Identity verified REQUIRED STEP — Agent Script AI handles response Verification tool call must return confirmed — no bypass ✓ Deterministic: required steps always execute AI reasoning operates within defined rails The problem Agent Script was built to solve AI agents are probabilistic systems. Give an agent a task and it reasons toward an outcome — evaluating options, choosing actions, interpreting ambiguous inputs — in a way that is flexible and often impressive but never fully predictable from the instruction set alone. For customer service, sales qualification, and internal help desk use cases, that flexibility is valuable. The agent can handle variations in how a question is asked, adapt to unexpected inputs, and reach correct outcomes through different reasoning paths. However, for enterprise deployments where specific steps must happen in a specific order — identity verification before account access, compliance disclosure before a financial recommendation, escalation to a human at a specific deal stage — probabilistic behaviour is a blocker, not a feature. The question is not whether the agent gets to the right answer. It is whether it follows the required path to get there. How Agent Script works Agent Script adds a deterministic layer on top of the AI reasoning layer. You define the fixed logic in a human-readable JSON expression language — the if/then rules, the required tool calls, the handoff triggers, the compliance checkpoints — and the LLM handles the conversational reasoning in between the defined steps. The mental model is a set of rails. Between the rails, the agent reasons freely and handles variation naturally. At the defined points on the rails, the behaviour is fixed: the agent must perform a specific action, verify a specific condition, or hand off to a specific resource before it can continue. Agent Script — identity verification before account access Simplified example // Step 1: Required identity verification // Agent cannot proceed until this returns “verified” { “step”: “verify_identity”, “type”: “required_tool_call”, “tool”: “IdentityVerificationService”, “condition”: { “result”: “verified” // must match — no bypass } } // Step 2: AI handles the response conversation // LLM reasoning operates freely here { “step”: “handle_account_query”, “type”: “llm_reasoning”, “context”: “account_data”, “guardrails”: [“no_pii_in_response”] } // Step 3: Handoff trigger if deal threshold met { “step”: “check_escalation”, “if”: { “deal_value”: { “$gt”: 50000 } }, “then”: { “action”: “handoff_to_human”, “queue”: “senior_sales_rep” } } Between the defined steps, the AI reasons freely. At the defined steps, the behaviour is fixed — the agent cannot proceed without completing the required action. The distinction from a traditional decision tree is important. Agent Script is not replacing AI reasoning with a flowchart. It is constraining the parts of the interaction where constraint is required while leaving the parts where AI is valuable free to operate as designed. Practical examples Customer service agent with identity verification Without Agent Script: The agent is instructed to verify identity before accessing account data. Whether it follows that instruction precisely depends on how the conversation goes — a persuasive or impatient customer might cause the agent to skip or abbreviate the step. With Agent Script: Identity verification is a defined step in the script. The agent cannot proceed to account lookup until the verification tool call returns a confirmed result. The LLM handles the conversation around verification, but the verification itself is not optional. Sales qualification agent with human handoff Without Agent Script: The agent is instructed to hand off to a human sales rep when a qualified opportunity reaches a specific threshold. Whether it recognises the threshold correctly depends on its interpretation of the conversation. With Agent Script: The handoff trigger is defined explicitly. When the deal value field exceeds the threshold, or when the prospect uses specific phrases indicating high purchase intent, the agent executes a defined handoff action to the Salesforce queue. The LLM does not decide whether the handoff happens; the script does. Integration with Agentforce Builder Agent Script integrates with Agentforce Builder — the conversational build environment introduced in Spring ’26. Admins and developers write Agent Script in the Builder interface alongside the natural language instructions that govern the agent’s conversational behaviour. The separation of concerns in the build environment mirrors the separation in the runtime: conversational behaviour is configured in natural language, deterministic controls are written in Agent Script. Neither replaces the other. Additionally, Agent Script was open-sourced at TDX 2026. The full specification, parser, and compiler are on GitHub. For compliance teams and legal teams who need to understand and audit agent behaviour independently of the vendor, this is significant: Agent Script is a readable, auditable format that does not require Salesforce expertise to understand the logic. Who needs Agent Script now, and who can wait Use case type Determinism non-negotiable Pure AI reasoning is fine Customer service — account access Identity verification must precede account lookup without exception. Agent Script enforces the required step regardless of conversation flow. Handling FAQ questions that do not involve personal data — product information, store hours, general policies. Financial services Regulatory disclosures must be delivered before a product recommendation. Compliance checkpoints cannot be skipped regardless of conversation length. Summarising publicly available market information with no personalised recommendation required. Sales qualification Handoff to a human rep must trigger at a specific deal stage or value threshold,
