Sf Ai Agentforce Testing

This is for teams running formal Agentforce agent tests after publish, not for building the agents themselves. It supports two tracks: multi-turn API testing when you need conversation continuity and context preservation, or CLI Testing Center for single-utterance validation and org-native test specs. The workflow is structured around a 100-point scoring system across seven categories, and it handles the full test-fix loop including topic routing validation, action invocation checks, and guardrail testing. If your test reveals the agent logic itself needs fixing, you hand off to the Agent Script skill. It requires API v66.0+ and assumes you are testing already-published agents, not prototypes.

Install to Claude Code

npx -y skills add jaganpro/sf-skills --skill sf-ai-agentforce-testing --agent claude-code