{"full_name":"SWE-agent/SWE-agent","owner":"SWE-agent","name":"SWE-agent","shard":"1","url":"https://github.com/SWE-agent/SWE-agent","description":"SWE-agent takes a GitHub issue and tries to automatically fix it, using your LM of choice. It can also be employed for offensive cybersecurity or competitive coding challenges. [NeurIPS 2024] ","homepage":"https://swe-agent.com","topics":["agent","ai","developer-tools","llm","agent-based-model","lms","cybersecurity"],"tag_pages":["agent","ai","cybersecurity","developer-tools","llm"],"language":"Python","license":"MIT","avatar":"https://avatars.githubusercontent.com/u/166046056?v=4","created_at":"2024-04-02T04:09:47Z","pushed_at":"2026-11-09T22:24:38Z","metrics":{"stars":20811,"forks":2270,"watchers":109,"issues":662,"prs":952,"releases":10,"issues_open":61,"issues_closed":601,"prs_open":128,"prs_closed":184,"prs_merged":640,"discussions":0},"growth":{"day":{"stars":7,"forks":2,"base_day":"2026-11-15"},"week":{"stars":95,"forks":6,"base_day":"2026-11-04"},"month":{"stars":233,"forks":-16,"base_day":"2026-10-17"}},"history":{"dates":["2026-09-06","2026-09-07","2026-09-08","2026-09-09","2026-09-10","2026-09-11","2026-09-12","2026-09-13","2026-09-14","2026-09-15","2026-09-16","2026-09-17","2026-09-18","2026-09-19","2026-09-20","2026-09-21","2026-09-22","2026-09-23","2026-09-24","2026-09-25","2026-09-26","2026-09-27","2026-09-28","2026-09-29","2026-09-30","2026-10-01","2026-10-02","2026-10-03","2026-10-04","2026-10-05","2026-10-06","2026-10-07","2026-10-08","2026-10-09","2026-10-10","2026-10-11","2026-10-12","2026-10-13","2026-10-14","2026-10-15","2026-10-16","2026-10-17","2026-10-18","2026-10-19","2026-10-20","2026-10-21","2026-10-22","2026-10-23","2026-10-24","2026-10-25","2026-10-26","2026-10-27","2026-10-28","2026-10-29","2026-10-30","2026-10-31","2026-11-01","2026-11-02","2026-11-03","2026-11-04","2026-11-10","2026-11-11","2026-11-12","2026-11-13","2026-11-14","2026-11-15","2026-11-16"],"stars":[20235,20256,20270,20285,20292,20297,20305,20309,20312,20326,20335,20340,20344,20355,20365,20370,20382,20384,20390,20396,20405,20416,20427,20436,20447,20448,20451,20464,20479,20488,20489,20495,20496,20508,20511,20516,20524,20536,20544,20554,20573,20578,20584,20594,20599,20603,20607,20614,20625,20632,20636,20643,20656,20670,20683,20684,20695,20703,20712,20716,20771,20780,20785,20787,20798,20804,20811],"forks":[2213,2214,2214,2213,2214,2217,2218,2218,2219,2218,2220,2221,2224,2225,2229,2230,2231,2232,2233,2234,2233,2236,2237,2239,2241,2244,2245,2248,2249,2248,2249,2249,2248,2247,2249,2248,2249,2252,2253,2257,2258,2286,2306,2324,2339,2330,2260,2259,2263,2263,2264,2266,2263,2259,2258,2259,2260,2261,2264,2264,2265,2266,2266,2267,2267,2268,2270],"watchers":[112,113,113,113,113,113,113,113,113,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,112,110,110,110,110,110,109,109,109,109,109,109,null,109,109,109,109,109,109],"issues_open":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,59,59,59,59,60,61],"prs_open":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,122,123,124,125,126,128]},"boards":{},"same_owner":[{"full_name":"SWE-agent/mini-swe-agent","name":"mini-swe-agent","shard":"1","boards":{}}],"issues":[{"number":1240,"title":"Buggy cost control with Gemini 2.5 Pro","comments":9,"created_at":"2025-07-10T02:48:28Z"},{"number":702,"title":"End-of-line-sequences: `\\r\\n` vs `\\n`","comments":7,"created_at":"2024-07-30T20:32:35Z"},{"number":1526,"title":"RFC: deterministic execution boundaries for tool calls (task text is untrusted input)","comments":5,"created_at":"2026-08-30T10:45:35Z"},{"number":1170,"title":"\"pip not found\" for tools/edit_anthropic when running SweBench MultiModal","comments":6,"created_at":"2025-05-27T23:32:49Z"},{"number":1333,"title":"How to correctly capture DeepSeek reasoning_content in agent trajectory?","comments":5,"created_at":"2026-01-22T03:55:13Z"},{"number":1302,"title":"Local model (Ollama CodeLlama 13B) + `thought_action` keeps emitting multi-line Python; SWE-agent treats it as bash and loops with `BashIncorrectSyntaxError`","comments":5,"created_at":"2025-09-18T05:13:36Z"},{"number":1051,"title":"SWE-agent for coding challenge fails to write patch due to Syntax Error","comments":5,"created_at":"2025-03-23T23:49:44Z"},{"number":1481,"title":"Cache-hit accounting is invisible in SWE-agent's own token stats","comments":3,"created_at":"2026-07-28T19:04:28Z"}],"discussions":[],"latest_release":{"name":"v1.1.0: 10s of thousands of training trajectories","tag":"v1.1.0","published_at":"2025-05-22T16:11:39Z","body":"# v1.1.0: 10s of thousands of training trajectories\r\n\r\nWe're very excited to announce our new project [SWE-smith](https://swesmith.com/), generating 10s of thousands of training trajectories for SWE agents.\r\nUsing this training data, our LM SWE-agent-LM-32b achieves open-weights SotA on SWE-bench verified with SWE-agent!\r\n\r\n![image](https://github.com/user-attachments/assets/e930ed68-de04-45e1-bd0d-776d87f202c0)\r\n\r\nApart from that, v1.1.0 is mostly a fix release with minor improvements, in particular adding compatibility with SWE-bench multilingual/multimodal, and SWE-smith. However, please pay attention to the breaking changes below.\r\n\r\n### Breaking changes\r\n\r\n* Changes to trajectory data format. The `messages` field is replaced by `query` by @klieret in https://github.com/SWE-agent/SWE-agent/pull/1107\r\n* Renamed many tool bundles that used \"windowed\" file viewer (`defaults` and more) by @klieret in https://github.com/SWE-agent/SWE-agent/pull/1147\r\n* Removed `review_on_submit` tool bundle (replaced by `review_on_submit_m`) by @klieret in https://github.com/SWE-agent/SWE-agent/pull/1148\r\n* Change in `windowed` tools (formerly `default`): Don't append \\n to new file by @klieret in https://github.com/SWE-agent/SWE-agent/pull/1114\r\n\r\n### Added\r\n\r\nNew dataset support:\r\n\r\n* Feat: Support multilingual evaluation by [@kabirgh](https://github.com/kabirgh) in [#1090](https://github.com/princeton-nlp/SWE-agent/pull/1090)\r\n* Feat: SWE-smith & multimodal base support by [@klieret](https://github.com/klieret) in [#1092](https://github.com/princeton-nlp/SWE-agent/pull/1092)\r\n\r\nNew utilities:\r\n\r\n* Feat: Add quick-stats tool by [@klieret](https://github.com/klieret) in [#1125](https://github.com/princeton-nlp/SWE-agent/pull/1125)\r\n\r\n### Enhanced\r\n\r\n* Feat: Config/override max_output_tokens by [@klieret](https://github.com/klieret) in [#1036](https://github.com/princeton-nlp/SWE-agent/pull/1036)\r\n* Enh: [#1042] fix(run_batch): handle JSON parsing errors in trajectory check by [@FRAOTIAC](https://github.com/FRAOTIAC) in [#1043](https://github.com/princeton-nlp/SWE-agent/pull/1043)\r\n* Enh: Allow to override tools dirs etc. by [@klieret](https://github.com/klieret) in [#1046](https://github.com/princeton-nlp/SWE-agent/pull/1046)\r\n* Enh: Allow to override path to swe-bench dataset by [@klieret](https://github.com/klieret) in [#1093](https://github.com/princeton-nlp/SWE-agent/pull/1093)\r\n* Enh: Allow to disable python-standalone for batch by [@klieret](https://github.com/klieret) in [#1115](https://github.com/princeton-nlp/SWE-agent/pull/1115)\r\n* Enh: More information on skipped exit status by [@klieret](https://github.com/klieret) in [#1117](https://github.com/princeton-nlp/SWE-agent/pull/1117)\r\n\r\n### Fixed\r\n\r\n* Fix: Setting max_input_tokens to 0 by [@klieret](https://github.com/klieret) in [#999](https://github.com/princeton-nlp/SWE-agent/pull/999)\r\n* Fix: Explicitly set log file encoding by [@klieret](https://github.com/klieret) in [#1013](https://github.com/princeton-nlp/SWE-agent/pull/1013)\r\n* Fix: Ensure pydantic-settings env prefix set by [@klieret](https://github.com/klieret) in [#1018](https://github.com/princeton-nlp/SWE-agent/pull/1018)\r\n* Fix: run batch processing with modal by [@vsee](https://github.com/vsee) in [#1023](https://github.com/princeton-nlp/SWE-agent/pull/1023)\r\n* Fix: Catch exit forfeit by [@klieret](https://github.com/klieret) in [#1024](https://github.com/princeton-nlp/SWE-agent/pull/1024)\r\n* Fix: Use 'latest' image tag for SWE-Bench images by [@klieret](https://github.com/klieret) in [#1029](https://github.com/princeton-nlp/SWE-agent/pull/1029)\r\n* Fix: Show tenacity retry reasons by [@klieret](https://github.com/klieret) in [#1032](https://github.com/princeton-nlp/SWE-agent/pull/1032)\r\n* Fix: Compatibility with textual 2.0 by [@klieret](https://github.com/klieret) in [#1033](https://github.com/princeton-nlp/SWE-agent/pull/1033)\r\n* Fix: Use default trajectories dir according to ENV by [@vsee](https://github.com/vsee) in ","is_prerelease":false},"readme":true,"readme_zh":false,"news":[],"stale":false,"snapshot_at":"2026-11-16T09:23:14Z","history_start":"2026-09-06"}