{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.10394", "kind": "paper", "url": "https://arxiv.org/abs/2606.10394"}], "authors": ["Sirui Liang", "Bohan Yu", "Peiyu Wang", "Shiguang Guo", "Wenxing Hu", "Pengfei Cao", "Jian Zhao", "Cao Liu", "Ke Zeng", "Xunliang Cai", "Kang Liu"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.10394", "source_url": "https://arxiv.org/abs/2606.10394"}, "categories": ["Agents", "Agents & Tool Use", "General AI"], "description": {"en": "STAGE-Claw is an automated framework for building and evaluating personal-agent tasks in state-based computing environments, with a benchmark of 40 tasks. Evaluations measure final system state correctness."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.10394", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.10394", "title": "STAGE-Claw"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["University of Chinese Academy of Sciences", "Chinese Academy of Social Sciences", "Shandong Institute of Automation", "Meituan (China)"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2606.10394", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2606.10394", "source_url": "https://openalex.org/W7164170691"}, "key": "claire-radar:2606.10394", "modality": null, "name": "STAGE-Claw", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "Medium", "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_stage-claw_61b30974", "origin_source": "arxiv", "recognition_confidence": "0.85", "record_sha256": "cb76b5fc2568354256a711a8d177ab0bc1492fcd1c5541b0c8887a343c88dc21", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.10394"}, "publisher": null, "released": "2026-06-09", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.10394", "source_url": "https://arxiv.org/abs/2606.10394"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-10394", "source": "claire_radar", "source_benchmark_id": "2606.10394", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Agents & Tool Use", "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": 0, "hfPaperUrl": "https://huggingface.co/papers/2606.10394", "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", "lastSuccessfulDate": "2026-09-24", "state": "stale"}}, "status": "stale"}, "capabilities": [], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "54d9942e93beef802a2cd9ebc1975ee378cf0b8dc7519b187f6a0826ebbafb06", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "STAGE-Claw is an automated framework for building and evaluating personal-agent tasks in state-based computing environments, with a benchmark of 40 tasks. Evaluations measure final system state correctness.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "evaluation protocol evidence"], "snippet": "Given a task hint, STAGE-Claw automatically creates and validates a realistic benchmark task with its environment, task prompts, ground truth, and related verification programs."}, "familyId": "bmf_002eae0d56e7", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_stage-claw_61b30974", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2606.10394", "paper": "https://arxiv.org/abs/2606.10394", "pdf": "https://arxiv.org/pdf/2606.10394", "project": null, "report": "https://arxiv.org/abs/2606.10394"}, "metrics": [], "motivation": "Large language models are increasingly used to power personal agents for everyday applications, but evaluating these agents remains a challenge.", "name": "STAGE-Claw", "oneLine": "STAGE-Claw is an automated framework for building and evaluating personal-agent tasks in state-based computing environments, with a benchmark of 40 tasks. Evaluations measure final system state correctness.", "paperTitle": "STAGE-Claw: Automated State-based Agent Benchmarking for Realistic Scenarios", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.10394", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {}, "readiness": "Paper only", "recognitionConfidence": 0.85, "relation": "introduces", "releasedAt": "2026-06-09", "source": {"authors": ["Sirui Liang", "Bohan Yu", "Peiyu Wang", "Shiguang Guo", "Wenxing Hu", "Pengfei Cao", "Jian Zhao", "Cao Liu", "Ke Zeng", "Xunliang Cai", "Kang Liu"], "categories": ["cs.AI"], "id": "2606.10394", "title": "STAGE-Claw: Automated State-based Agent Benchmarking for Realistic Scenarios", "type": "arxiv", "url": "https://arxiv.org/abs/2606.10394"}, "sourceUpdatedAt": "2026-06-09T04:16:35Z", "topics": ["Agents"], "venueAttempts": [], "whyItMatters": "The framework addresses the need for scalable and realistic evaluation of personal agents, moving beyond sandboxed and static tasks to state-based verification. This supports progress in agent reliability and practical deployment."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
