Skip to main content

Agentic test results

As AI-driven testing becomes a core part of modern quality engineering, ReportPortal provides first-class support for agentic test results — outcomes produced by autonomous AI agents rather than traditional scripted test automation.

Starting from version 26.0.3, ReportPortal can receive and display results from AI agentic testing pipelines alongside classic automation results. This gives teams a unified view of their entire testing landscape — human-authored tests, automated regression suites, and AI agent sessions — in one place.

Why report agentic results to ReportPortal?​

  • Full testing picture. Consolidate results from AI agents and traditional automation into a single dashboard to support confident release decisions.
  • Evidence-based verification. Capture logs, attachments, and step-level details from agent sessions so you can review exactly what each AI agent tested and why it passed or failed.
  • Traceability. Distinguish agentic runs from classic launches at a glance in the ReportPortal UI, making it easy to track coverage and trends over time.

How to report agentic results​

To integrate your AI testing workflows with ReportPortal, choose one of the following approaches.

Programmatically​

Use ReportPortal integration guides and reporting SDKs for various programming languages to send results from your agent pipeline via the ReportPortal API.

Using agent skills​

Coming soon.

In the meantime, you can explore and experiment with the goRP CLI skill to report results directly from an AI agent session.

Mark your launches as agentic

To distinguish agentic launches from classic automation in the ReportPortal UI, include the following system attribute when starting a launch:

{
  "key": "isAgentic",
  "value": true,
  "system": true
}

This attribute is hidden from end users but is used by ReportPortal to apply the agentic visual indicator on the Launches page along with extended filtering options.

Integration showcase​

The AgenticQA testing platform integrates with ReportPortal out of the box, streaming evidence from AI agent test sessions in real time.