Agentforce & AI articles
59 articles · AI agents
Agentforce GPT-4.1 Retirement: Move Agents to GPT-5.6 Luna
OpenAI retires GPT-4.1 near March 2027, so every Agentforce Builder agent still using it as its reasoning engine has to...
AiAgentDefinition: Migrate Agentforce from Bot at API 68.0
At API 68.0, Agentforce agents no longer retrieve or deploy as Bot and BotVersion, so pipelines still using those types...
Agentforce Prompt Injection: Hardening After SalesBleed
SalesBleed turned a Web-to-Lead form into a zero-click leak of Account data through Agentforce prompt injection. The...
Salesforce Koa: What a CRM Reasoning Model Changes in Your Org
Koa is Salesforce's first CRM reasoning model, post-trained on NVIDIA Nemotron 3 Super and announced at Dreamforce '26....
Agentforce Subagents: Split and Route an Overloaded Agent
A decomposition and routing guide for architects whose single Agentforce agent has outgrown itself. Covers start_agent,...
Salesforce Winter '27 Marketing Cloud Updates for Developers
Winter '27 brings a fair amount to Marketing Cloud for developers and admins: new Agentforce agents for marketing work,...
Salesforce AI: DigitalOcean Managed AI Path
DigitalOcean's managed AI and Functions as a Service (FaaS) are a credible alternative to Salesforce's native AI...
Private Connect for Secure AI with Agentforce
Private Connect gives Agentforce and Data 360 a private, audited network path into customer data, which is what gets AI...
Reliable Production AI: Durable Workflows
Getting AI from prototype to production is an orchestration problem as much as a prompting one. Durable workflows keep...
Salesforce Claude Integration: Developer Guide
Salesforce's deepening integration with Anthropic's Claude is shifting its AI strategy. Here is how Claude is becoming...
Salesforce Data Cloud: Programmatic Implementation with Claude
A programmatic run at Salesforce Data Cloud identity resolution with Claude Code and the Metadata Configuration...
Salesforce Data Cloud Implementation Challenges Explained
Salesforce Data Cloud is complex because of how it was layered together over time, blending CDP, data lakehouse, and...