Spinner logo QXQA

AXQAAI

Did You Know?


Dual Execution - Server & Smart Agent API Execution

Server & Smart Agent API Execution

This feature allows APIs inside a test case to be executed either on the AXQA server or locally via the Smart Agent, depending on the project’s network and security requirements.


Why it matters

  • Some APIs are only reachable from inside a private network

  • Server-based execution is not always suitable for internal or restricted environments

  • Teams need flexibility without changing test case logic


When to use it

  • When APIs are hosted behind internal firewalls

  • When testers must execute APIs locally on their machines

  • When the same test case must run in different environments


Core concepts

  • Execution Source – Defines where the test case runs (Server or Smart Agent)

  • Smart Agent – A local execution client connected securely to AXQA

  • Server Execution – Centralized execution managed by AXQA backend


How it works

  1. User selects the execution source for the test case

  2. AXQA decides where the API should run

  3. APIs are executed either on the server or via Smart Agent

  4. Results are collected and reported identically


How to use it

Step 1: Open the Test Case

Select the desired execution source (Server or Smart Agent).

Step 2: Assign the test case if using Smart Agent

Ensure the test case is assigned to a user with an active Smart Agent.

Step 3: Run the test case

Click Run - AXQA automatically routes execution to the correct environment.

Step 4: Review results

Results appear in the same UI regardless of where execution happened.


Best practices

  • Use Server execution for public or external APIs

  • Use Smart Agent for internal or restricted APIs

  • Keep test logic identical across execution sources


Common mistakes

❌ Trying to run internal APIs on the server
✔ Switch execution source to Smart Agent

❌ Assigning Smart Agent execution without an active agent
✔ Ensure the assigned user has a connected Smart Agent


Security & permissions

  • Execution source is enforced per test case

  • Smart Agent execution requires authenticated pairing

  • Unsafe endpoints are blocked in both modes


Related documentation

Last update: Jan. 30, 2026

Tools

A+ A-

Version

1