Show HN: Bugbusters.ai automated bugfixing using gpt-3
bugbusters.ai
We are just launching Bugbusters. Bugbusters is a GitHub-Bot that writes bugfixes for errors detected by application monitoring software like sentry. It examines an error's monitoring data, such as stack traces, in combination with the source code and git commit history to generate a potential fix. The bugfix will be submitted via a Pull Request that will also include details on the determined cause of the initial error and approach taken to fix it. You can simply create a GitHub Issue containing a link to the error in sentry and assign it to the Bugbusters-Bot.
In cases where the bot is unable to generate a solution, it assists a programmer in finding a fix by providing information and steps that could lead to a resolution. This may involve providing a list of potential error causes, online research results such as stack-overflow posts, as well as code changes (commits) that may have caused an error.
While we are currently focused on automated bug fixing, there are lots of other interesting features that we want to incorporate in the future, such as:
Avoiding the reoccurrence of a bug by generating unit tests or recommending steps such as a refactoring to avoid them in the future.
Integration into IDEs/terminals/std-err, which would allow developers to overcome errors during coding time much faster. A programmer could also be informed that their code has a certain error-proneness or that they are working on a critical section that has been causing errors in the past.
Improving PR-Reviews by tracking critical code sections and highlighting them during reviews. This would allow developers to consider previously caused bugs and improve the quality of their code.
Overall, we believe that Bugbusters has the potential to greatly improve the efficiency and effectiveness of bug fixing. Let us know what you think!