Hi community, Agentforce becomes much more valuable once it starts working with real customer conversations, emails, and support workflows. 

But that’s usually the point where teams begin restricting what the agents are actually allowed to see. 

The result is less context, limited workflows, and AI that ends up far less useful than originally expected. 

At InCountry, we built AgentCloak so Agentforce can work with cloaked/tokenized data while authorized users still see the original values inside Salesforce. 

That way, teams can keep the workflows fully usable without exposing sensitive customer data across the AI layer. Happy to share more if relevant.