You will learn how to lock SCAPI traffic to Managed Runtime with the SCAPI Identity Header, allow other integrations, harden SLAS, and rate limit AI crawlers using eCDN fields and rule expressions.
Enable SCAPI Request Identification per instance
Add a block rule for the instance's path
Place the rule after the Salesforce allow rule
Start in Log mode, then switch to Block
How to restrict Shopper API traffic to Managed Runtime, rolled out in stages
Bots increasingly call the B2C Commerce API directly, bypassing storefronts, and AI crawlers and shopping agents add search and LLM costs and skew analytics. The October 2026 release lets you manage the SCAPI zone in Business Manager and the CDN API, accept Shopper API traffic only from Managed Runtime through the new SCAPI Identity Header, and identify AI traffic on any eCDN zone with verified-bot fields. The SCAPI zone is provisioned automatically and shared across Development, Staging and Production, managed from Production. Some controls such as WAF managed rulesets, mTLS, TLS 1.3 settings and HSTS are not available.
To restrict traffic, enable the Managed Runtime SCAPI Request Identification feature per instance, then add a firewall rule that blocks Shopper API paths for that instance’s org-specific path segment, placed after the locked Salesforce allow rule. Start with Log, then switch to Block. Other integrations such as a custom BFF can be allowed with their own rules carrying a separate shared secret header that must stay server-side. For SLAS, Strict Client Auth requires a client secret header on login and authorize requests, though Composable Storefront and Storefront Next do not send it yet, so it should be enabled only for private clients called from your own server code. SLAS also throttles refresh-token exchanges to 3 per 60 seconds.
For AI traffic, the Known Bot and Verified Bot Category fields sort verified bots into 17 categories. The post recommends rate limiting AI Crawler and AI Search traffic on expensive endpoints such as product search rather than blocking it, with thresholds tuned from logs and Log mode run first. Spoofed or unverified automation needs behavioral detection from a bot management partner or Bot Management in front of eCDN.
You will learn a method to reconcile two CSV exports on a shared External ID, using spreadsheet formulas to flag records that are missing, extra, duplicated or have blank or mismatched keys.
Why read this
You will also see why to check parent objects first.
Export source and target with the shared key
Count blank and duplicate keys per sheet
Match each row in both directions
Review missing, extra and duplicate statuses
Repeat per object, parents first
Reconciling source and target exports by external key, in order
After a migration, record counts can look right while hiding records that never arrived, records loaded twice, and records with no source. The article argues that the Salesforce Id cannot be used for matching because it is regenerated on insert, so you need a key present on both sides, such as a legacy ID held in an External ID field like Legacy_ID__c.
The method uses two exports and a spreadsheet. Keys are first checked for blanks and duplicates with a case-sensitive EXACT-based count, since COUNTIF ignores case. A status formula on each sheet then labels each row as OK, Missing in target, Extra in target, Duplicate key, Blank key or similar, which is checked in both directions so that a duplicate in the target shows up only when both sides are compared.
The author recommends running this per object, starting with Accounts and then Contacts and Opportunities, because a missing parent explains gaps underneath it. The method does not compare field values or verify parent relationships; the article points to separate posts for those checks and mentions a paid tool it sells.
You will learn how MCPs, Skills and Plugins fit together in the AIforce extensibility stack and which new admin skills to pick up first.
Why read this
It gives a roadmap and the training trails to follow.
User triggers a Plugin
Skill fires
MCP executes
Output returns to user
How the AIforce building blocks work together when a user acts
The post introduces AIforce as a new layer of the AI extensibility stack and explains three building blocks. MCP, the Model Context Protocol, is a standard bridge between an AI interface and live systems, data and APIs, so admins can expose flows, APIs and Apex without writing custom code for each system. Skills define the exact steps, rules and formatting an AI needs to complete a task consistently, and Plugins package MCP connectors, Skills and interfaces into a single installable product. It describes a flow in which a user triggers a Plugin, a Skill fires, an MCP executes and the output returns to the user.
The author lists the new admin skills to develop: MCP management, including centralised org-wide connections and Agentic Identity; building deterministic agent logic with Super Agents and Subagents; lifecycle management of Plugins mapped to user roles; and using Agentforce Coworker, Agent Script, Builder Central and Agentforce Vibes for AI-assisted building. The roadmap moves from AI Trust, Security and Governance trails through hands-on work with Agentforce Studio to advanced building and packaging, then ongoing change management.
The article is largely an overview with links to trails, workshops and resources rather than technical implementation detail, so practitioners will get orientation more than hands-on specifics.
By Itai Asseo and Egor Pakhomov · Salesforce BlogYou will also see how a policy that treats updates as corrections differs from one that files them as new facts.
You will learn how Salesforce AI Research measured agent memory end to end and that a revised memory policy, not an architecture change, roughly doubled recall in Agentforce.
Why read this
You will also see how a policy that treats updates as corrections differs from one that files them as new facts.
Previous policy
Keeps less of what customers say
Files an update as an extra fact
Updated policy
Keeps more of what customers say
Treats an update as a correction
The updated memory policy compared with the one it replaced
Salesforce AI Research built an evaluation framework that generates long, multi-session conversations with specific facts planted along the way, runs them through the live Agentforce system, and then asks questions that depend on those facts. Testing the full path, from the moment a customer mentions something to the moment an agent acts on it, is presented as what makes the result trustworthy. The conversations come from ConvoMem, a public benchmark released in 2025.
The framework was used to improve Agentforce Memory, which became generally available in September. The updated memory policy holds on to more of what a customer says and treats an update as a correction to something the agent already knows, rather than as an additional fact stored alongside the first. According to the article, the policy recalls 1.8 times as many explicitly stated facts and 2.2 times as many facts that change over time compared with the policy it replaced. The underlying architecture is unchanged.
The authors argue that measuring in production matters because memory only behaves like memory under real conditions, and they cite a travel platform customer whose employee-facing agent now carries context across conversations. The article is largely promotional and gives few methodological details beyond the headline ratios.
You will learn how buyer accounts, buyer groups, entitlement policies and price books control what each B2B customer sees and pays, and which questions to settle before connecting the storefront to an ERP.
You will learn how to write an HttpCalloutMock class, register it with Test.setMock, and cover error responses and request validation in Apex tests. It is a solid refresher, though the material is standard and widely documented.
You will learn a practical test for approval processes: check that requests reach a backup approver, that the backup has the margin data, and that the decision can be audited later. The piece also argues for agreeing approval rules with Sales and Finance before configuring anything.
You will see the reasoning behind setting agent capacity per channel, routing by skill, defining presence statuses and prioritising queues in Omni-Channel. The piece stays at the level of principles and gives no configuration steps.