Salesforce Expands Agent Fabric to Automatically Stop Runaway Costs and AI Agent Failures Before They Impact Customers
Salesforce, Tuesday, October 27th, 2026
Salesforce adds Agent Evals, API Evals, and cost controls to Agent Fabric to catch AI agent failures and cap spending.
Salesforce expanded Agent Fabric, its control plane for managing AI agents across platforms.
Agent Evals continuously tests agents, LLMs, and MCP servers in production and can trigger actions such as pulling an agent offline, while API Evals generates tests from API specifications to catch breaking changes.
A FinOps Dashboard and Model Wallet give a single view of AI spending with hard budget limits, and Cost Optimization Policies cut token usage by up to 75% in tests.
A Custom Agent Scanner and policy management for Azure, Apigee, and Kong gateways are also added, with the Agent Kill Switch and cost controls generally available and the Evals entering beta in November 2026.