SWE-agent is SOTA on offensive cybersecurity

SWE-agent EnIGMA (Enhanced Interactive Generative Model Agent) is SOTA on offensive cybersecurity challenges, with a 3.3x improvement over previous agents on the NYU CTF challenge dataset. The EnIGMA project introduces multiple novelties that are available to all use cases of SWE-agent, such as Interactive Agent Tools and a Summarizer to handle long outputs.

Major additions

• Capability to run over CTF challenges
• Interactive Agent Tools, including gdb
• Summarizers to handle long outputs

Smaller additions

• Add filemap command in the spirit of repomap by @samuela in https://github.com/princeton-nlp/SWE-agent/pull/619
• Create config to run human eval style challenges by @ofirpress in https://github.com/princeton-nlp/SWE-agent/pull/658
• Add claude 3.5 sonnet to models by @carlosejimenez in https://github.com/princeton-nlp/SWE-agent/pull/601
• Enh: Warn if scrolling >= 3 times by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/626
• feat: support deepseek-coder LLM by @jcraftsman in https://github.com/princeton-nlp/SWE-agent/pull/638
• Enh: Make timeout for agent commands configurable by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/674
• Add support for new gpt-4o-mini model by @ivan4722 in https://github.com/princeton-nlp/SWE-agent/pull/693
• Groq Models Integration by @MohammedNagdy in https://github.com/princeton-nlp/SWE-agent/pull/721
• Make log level configurable; add TRACE level by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/612

Fixes

• Compatibility with SWE-bench 2.0 by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/671
• ensure variables work in special command docstring by @forresty in https://github.com/princeton-nlp/SWE-agent/pull/628
• Important fix: Catch CostLimitExceeded in retry because of format/block by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/682
• Fix: Handle empty traj in should_skip by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/616
• Fix for end-marker communicate: Exit status always 0/invalid by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/644
• Fix: Insufficient quoting of git commit message by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/646
• Fix nonsensical trajectory formatting for PRs by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/647
• Fix: sweunexpected keyword 'python_version' by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/692
• Fix: Use LONG_TIMEOUT for pre_install commands by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/695
• Fix: UnboundLocalError when catching decoding issue by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/709
• Also create empty patch files for completeness by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/725
• Fix: Raise ContextWindowExceeded instead of exit_cost by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/727
• Fix: Deal with non-utf8 encoded bytes in comm by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/731
• Fix: Handle spaces in repo names by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/734
• Fix: Ensure utils is part of package by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/742
• Fix: Submitting ' ' in human mode crashes container by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/749
• Fix: Block su as command by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/752
• Fix: SWE_AGENT_MODEL_MAX_RETRIES needs casting by @klieret in https://github.com/princeton-nlp/SWE-agent/pull/757

New Contributors

🎉 @talorabr, @udiboy1209, @haoranxi, @NickNameInvalid, @rollingcoconut joined the team to build EnIGMA 🎉

• @carlosejimenez made their first contribution in https://github.com/princeton-nlp/SWE-agent/pull/601
• @samefarrar made their first contribution in https://github.com/princeton-nlp/SWE-agent/pull/606
• @hubstrauss made their first contribution in https://github.com/princeton-nlp/SWE-agent/pull/625
• @samuela made their first contribution in https://github.com/princeton-nlp/SWE-agent/pull/619
• @forresty made their first contribution in https://github.com/princeton-nlp/SWE-agent/pull/628
• @jcraftsman made their first contribution in https://github.com/princeton-nlp/SWE-agent/pull/638
• @ivan4722 made their first contribution in https://github.com/princeton-nlp/SWE-agent/pull/693
• @JoshuaPurtell made their first contribution in https://github.com/princeton-nlp/SWE-agent/pull/703
• @MohammedNagdy made their first contribution in https://github.com/princeton-nlp/SWE-agent/pull/721
• @pdemro made their first contribution in https://github.com/princeton-nlp/SWE-agent/pull/729