10 Jun 2025

FailTale

Hello everyone,

I’m incredibly thrilled to share some exciting progress on a Proof of Concept I’ve been working on, called FailTale! 🐛 🔍

For anyone interested in the intersection of AI and test automation, you can check out the project directly on GitHub: FailTale on GitHub

The Challenge: Making Debugging Less of a Chore

We’ve all been there—a test fails in a complex CI pipeline, and the long, tedious process of debugging begins. Sifting through logs, trying to reproduce the issue, and identifying the root cause can be a significant drain on time and energy. The core idea behind FailTale is to streamline this process and get to the bottom of issues faster by leveraging the power of AI.

How FailTale Works

When a test fails, FailTale intelligently collects relevant context to build a complete picture of the failure. This includes:

  • Test steps being executed
  • The error stack trace
  • Logs and outputs from custom console commands
  • Screenshots of the state at the time of failure

Once this data is collected, it leverages Large Language Models (LLMs) to analyze everything and offer insights and hints about the potential root cause.

Early Success and What’s Next

I’ve already integrated FailTale in its first attempts into our Uyuni CI pipeline, and it’s incredibly motivating to see that it’s already starting to provide some genuinely interesting root cause hints for our test failures.

The work doesn’t stop here, of course. Here’s what I’m focused on now and what the future holds:

  • Enriching Context with a RAG System: I’m currently working on a Retrieval-Augmented Generation (RAG) system. The goal is to process the extensive Uyuni documentation to further enrich the context provided to the AI, which should lead to even more accurate insights. It’s a bit more complex, which also makes it a bit more fun!

  • Future Architecture (Decoupling): Looking ahead, I’m considering decoupling the execution of commands (which connect to components via SSH) into a new, dedicated MCP Server. This would make the architecture even more robust and modular.

Join the Journey!

I’m really excited about where this project could go and hope to share a more detailed demo or perhaps even a talk about it later this year as it evolves.

For anyone working with test automation, especially within a Cucumber Test Framework, I’d love for you to take a look at the project on GitHub. Feedback, ideas, contributions, or even just a star would be fantastic. Let’s make debugging less of a chore, together!


Tags:
0 comments