The latest release of swe-agent/swe-agent is v1.1.0. This repository is an agent that takes GitHub issues and tries to resolve them automatically with a language model chosen by the user. The README states that SWE-agent has been replaced by mini-SWE-agent and recommends using mini-SWE-agent going forward.
The selected model autonomously uses tools to fix issues in real GitHub repositories, and can find cybersecurity vulnerabilities or perform custom tasks. The configuration is organized into a single yaml file and has been made concise and easy to modify for research use. The repository description says this agent can also be used for competitive coding challenges or offensive cybersecurity work. The language is Python, the license is MIT, the star count is 20605, and it was created on April 2, 2024. The repository description lists NeurIPS 2024.
The source of the confirmed changes is the README, not the release notes. There is no release date and no list of changes for v1.1.0, so it is not possible to know what that version fixed. No performance benchmarks are given either. The research record's momentum score remains low at 3.0 out of 10, with no freshness and no technology/quality scores. Since the README recommends using mini-SWE-agent, there appears to be less reason to keep tracking SWE-agent itself. If new features are what you are after, it is better to check mini-SWE-agent first.