{"record": {"aliases": ["An Empirical Study of Coordination Mode as the First-Class Citizen in From-Scratch Multi-Agent Coding"], "artifacts": [{"id": "arxiv:2607.27877", "kind": "paper", "url": "https://arxiv.org/abs/2607.27877"}, {"id": "gh:robinren03/mseval", "kind": "repo", "url": "https://github.com/robinren03/MSEval"}], "authors": ["Yanyu Ren", "Yunfeng Bai", "Xizheng Wang", "Li Chen", "Dan Li"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.27877", "source_url": "https://arxiv.org/abs/2607.27877"}, "categories": ["Agents", "General AI", "Language & Knowledge"], "description": {"en": "MSEval evaluates multi-agent software development on authentic full-stack projects using deterministic, requirement-based grading."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.27877", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.27877", "title": "MSEval"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Tsinghua University"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2607.27877", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2607.27877", "source_url": "https://openalex.org/W7171961290"}, "key": "claire-radar:2607.27877", "modality": null, "name": "MSEval", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "confidence": "Medium", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-09-25T04:40:22Z", "data_status": "primary-source-indexed", "display_eligible": "true", "origin_record_id": "bm_an-empirical-study-of-coordination-mode-as_3d405bb1", "origin_source": "arxiv", "recognition_confidence": "0.7", "record_sha256": "9001377b782451da051949ef5105275ebc61065bcaa4dbd5fb0229793b4bf6f5", "relation": "introduces", "review_model": "codex-cli-default", "review_state": "ai-reviewed", "reviewed_at": "2026-09-24T17:33:12.802650Z", "source_url": "https://arxiv.org/abs/2607.27877"}, "publisher": null, "released": "2026-07-30", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.27877", "source_url": "https://arxiv.org/abs/2607.27877"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-27877", "source": "claire_radar", "source_benchmark_id": "2607.27877", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["An Empirical Study of Coordination Mode as the First-Class Citizen in From-Scratch Multi-Agent Coding"], "annotation": "Machine Generated", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The combination of multi-agent coding, real-world projects, cost-aware measurement, and a public release is likely to draw near-term attention.", "score": 76}, "capabilities": [], "confidence": "Medium", "construction": "Original Synthetic", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-24T17:33:12.802650Z", "inputHash": "045fb148f34a85e5a1bf7163b1cace720d02972d60c32cac0b21eddbf9490aaa", "model": "codex-cli-default"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "We introduce multi-agent from-scratch evaluation benchmark, MSEval", "canonicalNameSource": "abstract", "decisionReason": "The supplied text provides a formal benchmark name, a repeatable task suite, deterministic scoring, and a public benchmark repository.", "evaluationEvidence": {"reuse": "The benchmark is released at https://github.com/robinren03/MSEval.", "scoring": "MSEval scores performance using hierarchical requirements and deterministic rubrics.", "task": "evaluating multi-agent coding on real-world tasks"}, "model": "codex-cli-default", "reviewedAt": "2026-09-24T17:33:12.802650Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "MSEval evaluates multi-agent software development on authentic full-stack projects using deterministic, requirement-based grading.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL"], "snippet": "We introduce multi-agent from-scratch evaluation benchmark, MSEval, evaluating multi-agent coding on real-world tasks."}, "familyId": "bmf_c69c7bd024c7", "firstSeenAt": "2026-08-19", "heat": null, "id": "bm_an-empirical-study-of-coordination-mode-as_3d405bb1", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/robinren03/MSEval", "data": null, "paper": "https://arxiv.org/abs/2607.27877", "pdf": "https://arxiv.org/pdf/2607.27877", "project": null, "report": "https://arxiv.org/abs/2607.27877"}, "metrics": [], "motivation": "Multi-agent vibe coding promises to accelerate software development, yet existing benchmarks rely on synthetic environments that ignore practical time and monetary costs, conflate reasoning with communication, and reward only superficial completion.", "name": "MSEval", "oneLine": "MSEval evaluates multi-agent software development on authentic full-stack projects using deterministic, requirement-based grading.", "paperTitle": "An Empirical Study of Coordination Mode as the First-Class Citizen in From-Scratch Multi-Agent Coding", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.27877", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-08-19T11:05:13.395391Z"}, "publications": [], "readiness": "Runnable", "recognitionConfidence": 0.7, "relation": "introduces", "releasedAt": "2026-07-30", "reviewLifecycle": {"state": "pending"}, "source": {"authors": ["Yanyu Ren", "Yunfeng Bai", "Xizheng Wang", "Li Chen", "Dan Li"], "categories": ["cs.AI", "cs.MA"], "id": "2607.27877", "title": "An Empirical Study of Coordination Mode as the First-Class Citizen in From-Scratch Multi-Agent Coding", "type": "arxiv", "url": "https://arxiv.org/abs/2607.27877"}, "sourceUpdatedAt": "2026-07-30T08:53:05Z", "topics": ["Agents"], "venueAttempts": [], "whyItMatters": "It supports comparable evaluation of functional quality, latency, and token cost across collaboration topologies."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
