{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2607.02606", "kind": "paper", "url": "https://arxiv.org/abs/2607.02606"}], "authors": ["Qirui Jin", "Lingching Tung", "Kenan Li", "Qiyang Shi", "Yushi She", "Huanzhong Jia", "Harrison Zhao", "Kejing Xia", "Zhenbang Du", "Jiaxin Pei", "Zhenyu Zhang", "Zhen Qi", "Yuyan Duan", "Wenke Lee", "Zijian Jin"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.02606", "source_url": "https://arxiv.org/abs/2607.02606"}, "categories": ["Code & Software", "Software & AI Compute", "cs.SE"], "description": {"en": "ChainSWE evaluates coding agents on sequential, dependent bug fixes within a shared codebase. It includes 304 issues across 54 Python projects, forming chronological chains, and measures performance drop as chain length increases."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.02606", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.02606", "title": "ChainSWE"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2607.02606", "modality": null, "name": "ChainSWE", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "unknown", "origin_record_id": "bm_chainswe_b0efa40a", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "9816617fec0cade2931387237149510cf1a7da5fe43eb112536e7d09a137227e", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2607.02606"}, "publisher": null, "released": "2026-07-01", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.02606", "source_url": "https://arxiv.org/abs/2607.02606"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-02606", "source": "claire_radar", "source_benchmark_id": "2607.02606", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["Software & AI Compute"], "area": "Code & Software", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2607.02606", "observedAt": "2026-10-04T03:02:42.754940Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "state": "unavailable"}}, "status": "unavailable"}, "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "57a51864fdbc645474213d93abaa4481ad903fabf6bc64fef5231455c5c3f4d1", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "ChainSWE evaluates coding agents on sequential, dependent bug fixes within a shared codebase. It includes 304 issues across 54 Python projects, forming chronological chains, and measures performance drop as chain length increases.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence"], "snippet": "To bridge this gap, we introduce ChainSWE, the first benchmark for evaluating agents on sequential, dependent bug fixes within a shared codebase."}, "familyId": "bmf_9b5273a86b4c", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_chainswe_b0efa40a", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Software & Cloud"], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2607.02606", "paper": "https://arxiv.org/abs/2607.02606", "pdf": "https://arxiv.org/pdf/2607.02606", "project": null, "report": "https://arxiv.org/abs/2607.02606"}, "metrics": [], "motivation": "Language model (LM) agents are increasingly deployed to maintain codebases over extended periods, fixing streams of related defects while carrying context from one fix to the next.", "name": "ChainSWE", "oneLine": "ChainSWE evaluates coding agents on sequential, dependent bug fixes within a shared codebase. It includes 304 issues across 54 Python projects, forming chronological chains, and measures performance drop as chain length increases.", "paperTitle": "ChainSWE: Benchmarking Coding Agents on Multi-Bug Software Maintenance", "primaryDomain": "Software & AI Compute", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.02606", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-07-01", "source": {"authors": ["Qirui Jin", "Lingching Tung", "Kenan Li", "Qiyang Shi", "Yushi She", "Huanzhong Jia", "Harrison Zhao", "Kejing Xia", "Zhenbang Du", "Yikai Zhang", "Jiaxin Pei", "Zhenyu Zhang", "Zhen Qi", "Yuyan Duan", "Wenke Lee", "Zijian Jin"], "categories": ["cs.SE"], "id": "2607.02606", "title": "ChainSWE: Benchmarking Coding Agents on Multi-Bug Software Maintenance", "type": "arxiv", "url": "https://arxiv.org/abs/2607.02606"}, "sourceUpdatedAt": "2026-07-01T13:32:15Z", "topics": ["cs.SE"], "venueAttempts": [], "whyItMatters": "Real-world software maintenance involves streams of related defects, but existing benchmarks evaluate one bug at a time. ChainSWE fills this gap by benchmarking agents on continuous workflows, revealing significant performance degradation on longer chains."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
