SWE-agent takes a GitHub issue and tries to automatically fix it, using GPT-4, or your LM of choice. It can also be employed for offensive cybersecurity or competitive coding challenges.
The --verbose flag isn't great, let's just set logging.setLevel numbers. Also, since it's not distinguishable from the CLI, let's make verbose affect all logs, rather than just the env.
Should we have a debug level where you can print all commands issued in the env and their output?
The
--verboseflag isn't great, let's just setlogging.setLevelnumbers. Also, since it's not distinguishable from the CLI, let's makeverboseaffect all logs, rather than just the env.Should we have a debug level where you can print all commands issued in the env and their output?