Changelog

Simulate Tools: Test Your Agent Without Running Real Tools

You can now test your agent's decision-making and tool usage without triggering actual API calls or incurring costs!

The new Simulate toggle lets you safely test your agent's behavior with mock tool outputs. This feature creates a sandbox environment where you can validate your agent's logic and tool selection without impacting production systems or using expensive API credits.

➡️ Test without consequences – Run your agent in a safe environment with no real API calls

➡️ Mock tool outputs – See how your agent handles different response scenarios

➡️ Validate decision trees – Ensure your agent selects the right tools at the right time

➡️ Test incomplete tools – Work with tools that aren't fully configured yet

➡️ Preserve API credits – Avoid unnecessary costs during development and testing

With Simulate Tools, you can confidently develop and refine your agent's workflows, ensuring they're production-ready before deploying them with real tool connections.

To use this feature, go to edit your agent under 'Build' > 'Run Test' > and simply toggle the "Simulate" option for select tools. In contrast the "Default" option will run the tool. Give your agent a test task, and the system will automatically generate mock responses for the select simulated tools instead of executing actual tool calls.

Start testing your agent's tool usage safely today!