Agent Testing

Build confidence in your agent's behavior with automated testing

The agent testing framework enables you to move from slow, manual phone calls to a fast, automated, and repeatable testing process. Create comprehensive test suites that verify both conversational responses and tool usage, ensuring your agents behave exactly as intended before deploying to production.

Video Walkthrough

Overview

The framework consists of two complementary testing approaches:

Scenario Testing (LLM Evaluation) - Validates conversational abilities and response quality
Tool Call Testing - Ensures proper tool usage and parameter validation
Simulation Testing - Runs end-to-end, multi-turn conversations with a simulated user

Both test types can be created from scratch or directly from existing conversations, allowing you to quickly turn real-world interactions into repeatable test cases.

Scenario Testing (LLM Evaluation)

Scenario testing evaluates your agent’s conversational abilities by simulating interactions and assessing responses against defined success criteria.

Creating a Scenario Test

Define the scenario

Create context for the text. This can be multiple turns of interaction that sets up the specific scenario you want to evaluate. Our testing framework currently only supports evaluating a single next step in the conversation. For simulating entire conversations, see our simulate conversation endpoint and conversation simulation guide.

Example scenario:

User: "I'd like to cancel my subscription. I've been charged twice this month and I'm frustrated."

Set success criteria

Describe in plain language what the agent’s response should achieve. Be specific about the expected behavior, tone, and actions.

Example criteria:

The agent should acknowledge the customer’s frustration with empathy
The agent should offer to investigate the duplicate charge
The agent should provide clear next steps for cancellation or resolution
The agent should maintain a professional and helpful tone

Provide examples

Supply both success and failure examples to help the evaluator understand the nuances of your criteria.

Success Example:

“I understand how frustrating duplicate charges can be. Let me look into this right away for you. I can see there were indeed two charges this month - I’ll process a refund for the duplicate charge immediately. Would you still like to proceed with cancellation, or would you prefer to continue once this is resolved?”

Failure Example:

“You need to contact billing department for refund issues. Your subscription will be cancelled.”

Run the test

Execute the test to simulate the conversation with your agent. An LLM evaluator compares the actual response against your success criteria and examples to determine pass/fail status.

Creating Tests from Conversations

Transform real conversations into test cases with a single click. This powerful feature creates a feedback loop for continuous improvement based on actual performance.

When reviewing call history, if you identify a conversation where the agent didn’t perform well:

Click “Create test from this conversation”
The framework automatically populates the scenario with the actual conversation context
Define what the correct behavior should have been
Add the test to your suite to prevent similar issues in the future

Tool Call Testing

Tool call testing verifies that your agent correctly uses tools and passes the right parameters in specific situations. This is critical for actions like call transfers, data lookups, or external integrations.

Creating a Tool Call Test

Select the tool

Choose which tool you expect the agent to call in the given scenario (e.g., transfer_to_number, end_call, lookup_order).

Define expected parameters

Specify what data the agent should pass to the tool. You have three validation methods:

Validation Methods

Exact Match
The parameter must exactly match your specified value.

Transfer number: +447771117777

Regex Pattern The parameter must match a specific pattern.

Order ID: ^ORD-[0-9]{8}$

LLM Evaluation An LLM evaluates if the parameter is semantically correct based on context.

Message: "Should be a polite message mentioning the connection"

Configure dynamic variables

When testing in development, use dynamic variable values that match those that would be actual values in production. Example: {{ customer_name }} or {{ order_id }}

Run and validate

Execute the test to ensure the agent calls the correct tool with proper parameters.

Critical Use Cases

Tool call testing is essential for high-stakes scenarios:

Emergency Transfers: Ensure medical emergencies always route to the correct number
Data Security: Verify sensitive information is never passed to unauthorized tools
Business Logic: Confirm order lookups use valid formats and authentication

Simulation Testing

Simulation testing evaluates your agent across a full, multi-turn conversation with a simulated AI user. Unlike single-turn evaluations, this test type checks whether the complete interaction reaches your defined outcome.

Simulation testing is currently in public alpha. Functionality and UI behavior may change.

Creating a Simulation Test

Define the simulation scenario

Describe the user’s context, intent, and behavior in natural language. The simulator uses this scenario to drive the conversation.

Example scenario:

“A tourist who is not fluent in English is trying to place an order at a restaurant.”

Set the success condition

Define the outcome that should count as a pass. This prompt is used to evaluate whether the full conversation succeeded.

Example success condition:

“The agent confirmed the order details, handled clarifying questions, and completed the order without misunderstandings.”

Set max turns

Choose how long the simulation can run before stopping. Use a lower value for focused checks and a higher value for complex workflows.

Minimum: 1
Maximum: 50
Default: 5

Run and review the result

Execute the test and inspect the generated conversation transcript. Review the pass/fail result against your success condition, then iterate on your prompt, tools, or agent configuration.

Optional Configuration

You can refine simulation behavior in the test configuration panel:

Environment: Select which environment to test against when your agent has multiple environments configured. If only one environment is available, this selector is hidden.
Chat history: Start from a partial conversation instead of a blank state. This is useful for testing in-progress conversations and recovery behavior.
Dynamic variables: Inject test-specific values into your agent variables (for example, user names or order IDs) without changing the base agent configuration.

Tool Mocking

Simulation tests support tool mocking so your agent can receive controlled responses during a run instead of calling live systems.

Mocking strategy

Mock none: No tools are mocked.
Mock all tools: Every mockable tool returns a mock response.
Mock selected tools: Only tools you explicitly choose are mocked.

System tools and workflow tools are never mocked.

Fallback behavior

If a mocked tool is called and no matching mock response is found, choose one of these behaviors:

Call real tool: Executes the real tool call.
Finish with error: Returns an error response from the tool instead of calling the real tool.

The fallback setting appears only when at least one tool is mocked.

Development Workflow

The framework supports an iterative development cycle that accelerates agent refinement:

Write tests first

Define the desired behavior by creating tests for new features or identified issues.

Test and iterate

Run tests instantly without saving changes. Watch them fail, then adjust your agent’s prompts or configuration.

Refine until passing

Continue tweaking and re-running tests until all pass. The framework provides immediate feedback without requiring deployment.

Save with confidence

Once tests pass, save your changes knowing the agent behaves as intended.

Running Tests

Run via the dashboard

Run via the CLI

Run via the API

Navigate to the Tests tab in your agent’s interface. From there, you can run individual tests or execute your entire test suite at once using the “Run All Tests” button.

Batch Testing and CI/CD Integration

Execute all tests at once to ensure comprehensive coverage:

Select multiple tests from your test library
Run as a batch to identify any regressions
Review consolidated results showing pass/fail status for each test

Best Practices

Evaluate agent persona consistency

Test that your agent maintains its defined personality, tone, and behavioral boundaries across diverse conversation scenarios and emotional contexts.

Verify complex multi-turn reasoning

Create scenarios that test the agent’s ability to maintain context, follow conditional logic, and handle state transitions across extended conversations.

Test against prompt injection attempts

Evaluate how your agent responds to attempts to override its instructions or extract sensitive system information through adversarial inputs.

Assess ambiguous intent resolution

Test how effectively your agent clarifies vague requests, handles conflicting information, and navigates situations where user intent is unclear.

Next Steps

View CLI Documentation for automated testing setup
Explore Tool Configuration to understand available tools
Read the Prompting Guide for writing testable prompts