What Is Log Detective and Why It Matters for Packit Users
If you’ve ever scratched your head over a mysterious package build failure in Packit, relief is on the way. Starting this month, Log Detective integrates directly with Packit to automatically analyze failed scratch Koji builds triggered by dist-git pull requests. Packit already excels at bridging upstream projects with downstream distributions; now it gains an AI assistant that explains why a build broke—no extra clicks or configuration required.
How Log Detective Works Inside Packit
The integration is seamless. Whenever Packit triggers a Koji build that fails, it automatically requests an analysis from Log Detective. The user does nothing—no log selection, no prompt tuning. The analysis runs in the background and the result appears in the Packit dashboard as soon as it’s ready.
Log Parsing and Analysis Derivation
Under the hood, Log Detective (version 4.0 and later) is an agent built on the BeeAI Framework. When it receives logs and build artifacts, it employs a suite of tools based on the Drain template mining algorithm to extract only the most relevant snippets. These snippets represent a tiny fraction of the original log file size, which means:
- Fewer tokens consumed – The analysis finishes faster and costs less.
- Cleaner model context – Irrelevant noise is stripped away, improving accuracy.
- Smaller models can be used – Even relatively lightweight AI models yield good results on focused snippets.
Communication Architecture
Packit continues to handle failed Koji builds exactly as before. The difference is a new step: Packit sends a request to the Log Detective interface server—a lightweight, containerized service that manages all communication between the two systems. Once the analysis is complete, the interface server publishes the results on the Fedora Messaging bus, where Packit picks them up. This decoupled design ensures neither service blocks the other.
What You See in the Packit Dashboard
The Log Detective analysis consists of two parts:
- An explanation of what went wrong during the build (if anything was detected).
- An optional suggestion for how to fix it.
Currently, the analysis relies solely on the build logs—it does not consult external sources like upstream documentation or previous builds. The result is linked directly to the pull request that triggered the analysis, so developers can find it immediately in the Packit dashboard.
Purpose and Limitations of Log Detective
Log Detective uses a general-purpose AI model with no access to additional context. That means it cannot replace the deep expertise of experienced Fedora package maintainers. If you’ve been building packages for years, this tool probably won’t teach you anything new.
Instead, Log Detective is designed to help newcomers—developers who are less familiar with the Fedora ecosystem or with build troubleshooting. By providing a plain-language explanation and a suggested fix, it lowers the barrier to entry and makes package maintenance more accessible.
What’s Next for Log Detective
The engineering team has plans to extend Log Detective’s capabilities. Future versions may incorporate upstream repository analysis and other data sources, moving beyond purely log-based deductions. However, the core mission remains: to assist, not replace, human maintainers.
Log Detective for Packit is available now. No configuration is needed—just trigger a build and let the AI do the detective work.