testing-agentforce

Automate Agentforce test suites by using YAML test specs to plan, execute, and analyze end-to-end agent tests within CI/CD pipelines. You can deploy, run batch tests, and iterate using parsed pass/fail results and failure diagnostics.

End-to-end testing for Agentforce agents validates routing, actions, and outcomes across multi-turn conversations and guardrails. It works by auto-deriving test plans from agent definitions and executing structured YAML test suites to parse pass/fail results.

Yes, you can run batch tests in Testing Center using YAML specs. The skill supports Mode A/B testing, allowing you to execute ad-hoc previews or batch tests and parse the results for clear failure diagnostics and reporting.

Validate Agentforce guardrails and safety guidelines by incorporating safety probes into your auto-derived test plans. The skill executes structured YAML test suites and parses results to ensure agents adhere to defined safety guardrails during conversations.

Yes, you need YAML files to author and validate AiEvaluationDefinition-like test specs. The skill applies to workflows where test specs are structured in YAML, enabling you to plan, execute, and analyze structured test suites for your agents.

Agentforce sf agent test commands may fail due to routing, action, or guardrail violations detected during multi-turn conversations. The skill parses test results to provide clear failure diagnostics, enabling you to iterate and fix issues within your CI/CD workflows.