Test Scenario Creation Agent

​

What It Does

• Understands Your APIs Deeply Parses your uploaded Swagger/OpenAPI/Postman collection, builds a dependency graph, and identifies how endpoints interact.
• Generates End-to-End Test Flows You can prompt in natural language (e.g., “Write positive login + user create + fetch scenario”) and the agent automatically:
  • Selects relevant APIs
  • Builds a logical step-by-step flow
  • Chains data between calls
  • Adds placeholder or AI-generated input data
• Supports Customization Easily add, remove, or edit steps. You can toggle between AI-generated and manual mode.
• Includes Modular Scenario Suggestions If your APIs are tagged by module (e.g., auth, users, transactions), you can generate test scenarios per module or across them.

​

How It Works

1. Go to the AI Agent interface → Launch here
2. Type what you want to test Example prompts:
   • “Test login with valid and invalid credentials”
   • “Check user creation without email”
   • “Verify if access is denied without token”
3. AI Agent will respond with suggestions You’ll see proposed test scenarios and coverage types (positive, negative, auth, etc.).
4. Click “Save” to confirm the test This will auto-create:
   • A new Test Scenario
   • One or more Test Cases inside it
   • Draft Test Rules (assertions, status codes, auth checks)
   • Support for API chaining
5. Preview or modify the logic Click “Code” to see and edit the YAML/JS version of the test.
6. Generate request data or rules if needed After saving, use the “Create Request Data” or rule options to extend or fine-tune tests.

​

Need to write scenarios manually?

​

Test Scenario Creation — FAQs

1. How does the AI know which APIs to include in a scenario? Qodex builds a dependency graph from your collection and uses your prompt to find a logical chain of APIs based on auth flows, object creation, and response dependencies.
2. Can I write my own test scenario from scratch? Yes. While the agent speeds up the process, you can switch to manual mode and drag/drop APIs, set chaining manually, or import your own flow.
3. What kind of prompts work best? Prompts should be concise but clear. Use keywords like:
   • Positive/negative
   • Auth/login/refresh
   • Chained flow/user journey
   • OWASP/security/payload abuse
   • Specific module names or API paths
   Example: “Positive checkout with new user, add item, apply coupon, pay”
4. How many scenarios can I generate in a project? There’s no upper limit, but token usage applies. For large projects, it’s best to generate test scenarios module-wise or in logical flows (e.g., auth first, then business logic).
5. Can I collaborate with my team on scenarios? Yes. Your workspace allows team members to view, edit, and run shared test scenarios. You can also sync everything to GitHub for full visibility.
6. Do I need to write any code? No. The agent handles everything from flow building to rule writing. If you prefer, you can view/edit the generated code, but it’s fully no-code by default.