Today4 stories

The Daily Org

Spring '27Sandboxes 6 Feb 2027

Curated by Abhinav (opens in a new tab)

Vol. I · No. 10The independent daily for Salesforce admins, developers and architects4 stories

Must-readAgentforce and AI

How Agentforce Coworker delegates requests to Employee Agents (opens in a new tab)

You will learn how Agentforce Coworker selects which agent handles a request and why only agent descriptions, not subagent instructions, control that decision.

Why read this

Readers can apply these rules to configure reliable multi-agent routing and avoid common delegation pitfalls.

  1. User submits a request to Agentforce Coworker
  2. Coworker cannot complete the request itself
  3. Reads accessible agent and subagent descriptions
  4. Compares request against agent capabilities
  5. Delegates request to matching Employee Agent
Agentforce Coworker routes incomplete requests by evaluating accessible agent descriptions against user queries

Agentforce Coworker acts as a central entry point that delegates incomplete requests to specialized agents like Agentforce Employee Agents. The system evaluates the incoming request against the capabilities defined in each accessible agent and subagent description before making a routing decision.

A critical configuration detail is that Coworker ignores subagent instructions when determining routing. Builders must place all routing-relevant information directly in the agent descriptions to ensure accurate matching. User permissions also filter which agents are available for delegation.

Reliable delegation requires distinct capability statements across agents, explicit coverage of subagents, and verification of user access levels. Testing with varied phrasing helps confirm that the routing logic aligns with actual user requests.

Read the original at InfallibleTechie (opens in a new tab)

Worth reading?

RecommendedReleases

Winter '27 release brings user history tracking and manual share retention (opens in a new tab)

Readers will learn how to enable native field history tracking on the User object, preserve manual record shares during ownership transfers, and leverage new Lightning App Builder layout controls without writing code.

Why read this

They will also discover diagnostic tools for monitoring Custom Field indexing and query frequency to reduce technical debt.

The Winter ’27 release makes field history tracking generally available for the User object, allowing administrators to monitor up to twenty fields such as profile, role, and active status directly through the Object Manager. An org-wide toggle under Sharing Settings now prevents manual record shares from being deleted when a record owner changes, which helps maintain collaboration during territory restructures.

The update expands Lightning App Builder capabilities by adding dynamic sorting for Related Lists based on formulas or lookups, conditional section headers that show or hide accordion groups, and scheduling tools for in-app prompts to prevent user fatigue. These layout enhancements align with SLDS 2 Cosmos accessibility standards and require no custom components.

Administrators can now view index status reports for Custom Fields within the Performance Metrics dashboard to troubleshoot slow SOQL queries and report timeouts. Object Manager also displays last modified dates and query frequency for each Custom Field, enabling safer and faster cleanup of unused fields before deprecation.

Read the original at Salesforce Break (opens in a new tab)

Worth reading?

Architecture

Architecting customer portals beyond Experience Cloud licensing and API limits (opens in a new tab)

Readers will learn how to evaluate four portal architectures against Experience Cloud by weighing licensing costs, API budgets, and data synchronization trade-offs.

Why read this

The piece provides a structured set of questions to assess security, visibility, and failure modes for each approach.

The article examines four architectural approaches for building customer portals that do not rely on Experience Cloud licenses. It contrasts running the portal directly on the Salesforce platform using licensed external users or headless identity APIs, connecting to a live Salesforce database through a third party, syncing records to an external database, or migrating customer data entirely to a different system like Microsoft Dataverse.

Each model carries distinct operational trade-offs. On-platform deployments enforce native security and real-time data but require per-user licensing and consume administrative capacity. Off-platform live connections avoid licensing fees but draw heavily from the rolling 24-hour API limit and introduce OAuth token security risks. Synced copies reduce API consumption and provide outage resilience but risk data staleness and write collisions, requiring careful reconciliation policies.

To navigate these options, the author recommends evaluating every vendor against five criteria: where customer data resides, how Salesforce classifies portal visitors, which layer enforces record visibility, the typical API cost per page or acceptable sync lag, and procedures for handling connection failures or token revocation. The analysis concludes that Experience Cloud remains the only true single system of record, while alternatives simply shift architectural responsibilities to API management, data governance, or external security configurations.

Read the original at dev.to (opens in a new tab)

Worth reading?

Releases

B2C Commerce release 26.9 introduces edge credential detection and SFRA password warnings (opens in a new tab)

Readers will learn how to configure Cloudflare firewall and rate limiting rules in B2C Commerce to block leaked credentials, and how to implement server-side header checks in SFRA to prompt users to rotate breached passwords.

The September 2026 update to B2C Commerce adds built-in credential stuffing protection through its eCDN layer without requiring additional entitlements. Four new detection fields evaluate login attempts against known breach data, allowing administrators to create custom firewall and rate limiting rules at the network edge.

The article provides exact rule syntax examples to throttle malicious traffic while preserving verified search crawlers. It explains how to use managed challenges for password leaks and emphasizes separating counting expressions from rule expressions to avoid penalizing legitimate bot activity.

For storefront implementations, the piece details how to read a server-side request header added by Cloudflare after a successful login. A working SFRA code example demonstrates how to map header values to user-facing warning messages and direct shoppers to a password reset flow, deliberately avoiding failed login responses to prevent information leakage to attackers.

Finally, the author outlines monitoring strategies using Security Analytics in Business Manager and eCDN Logpush exports, recommending these detections as a baseline before layering in broader bot mitigation controls.

Read the original at Salesforce Developers Blog (opens in a new tab)

Worth reading?