{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2605.26086", "kind": "paper", "url": "https://arxiv.org/abs/2605.26086"}, {"id": "gh:libercoders/claw-anything", "kind": "repo", "url": "https://github.com/LiberCoders/Claw-Anything"}, {"id": "hf:libercoders/claw-anything", "kind": "dataset", "url": "https://huggingface.co/datasets/LiberCoders/Claw-Anything"}, {"kind": "website", "url": "https://libercoders.github.io/Claw-Anything/"}], "authors": ["Yusong Lin", "Xinyuan Liang", "Haiyang Wang", "Qipeng Gu", "Siqi Cheng", "Jiangui Chen", "Shuzhe Wu", "Feiyang Pan", "Lue Fan", "Sanyuan Zhao", "Dandan Tu"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2605.26086", "source_url": "https://arxiv.org/abs/2605.26086"}, "categories": ["Agents", "General AI", "Language & Knowledge"], "description": {"en": "Claw-Anything evaluates always-on LLM personal assistants in simulated environments with long-horizon activity histories, interdependent backend services, and GUI/CLI across devices. It includes 200 human-verified tasks scored on completion, robustness, communication, and safety, with a live leaderboard."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2605.26086", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2605.26086", "title": "Claw-Anything"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Beijing Institute of Technology", "Peking University", "Chinese Academy of Sciences", "Institute of Automation"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2605.26086", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2605.26086", "source_url": "https://openalex.org/W7162459613"}, "key": "claire-radar:2605.26086", "modality": null, "name": "Claw-Anything", "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-reviewed", "display_eligible": "unknown", "origin_record_id": "bm_claw-anything_3c56b6b4", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "6365d25f3c388f93bd597f7b6520c2e6e042d8d93d70877d5a3aa9053ce476ce", "relation": "introduces", "review_state": "source-reviewed", "reviewed_at": "2026-08-19", "source_url": "https://arxiv.org/abs/2605.26086"}, "publication_dates": [{"basis": "paper_first_version", "date": "2026-05-25", "source_url": "https://arxiv.org/abs/2605.26086"}], "publisher": null, "released": "2026-05-25", "released_reference": {"basis": "first_public", "source_key": "claire-radar:2605.26086", "source_url": "https://arxiv.org/abs/2605.26086"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2605-26086", "source": "claire_radar", "source_benchmark_id": "2605.26086", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/LiberCoders/Claw-Anything", "githubScope": "benchmark_repo", "githubStars": 48, "hfDailySubmittedAt": null, "hfDataset": "LiberCoders/Claw-Anything", "hfDatasetDownloads": 112, "hfDatasetLikes": 10, "hfPaperUpvotes": 26, "hfPaperUrl": "https://huggingface.co/papers/2605.26086", "observedAt": "2026-10-04T15:31:48.030916Z", "signalStatus": {"githubStars": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}, "hfDatasetDownloads": {"checkedAt": "2026-10-04T15:31:48.030916Z", "lastSuccessfulDate": "2026-09-30", "state": "stale"}, "hfDatasetLikes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "lastSuccessfulDate": "2026-09-30", "state": "stale"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "lastSuccessfulDate": "2026-09-30", "state": "stale"}}, "status": "fresh"}, "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": "a4ba1f383be74e2a945572f6c15d0cad22707c80dfa7a662d4ae181581c0c443", "model": "deepseek-v4-flash"}, "curation": {"reviewedAt": "2026-08-19", "sources": ["https://github.com/LiberCoders/Claw-Anything", "https://huggingface.co/datasets/LiberCoders/Claw-Anything", "https://arxiv.org/abs/2605.26086"], "state": "source-reviewed"}, "dataStatus": "primary-source-reviewed", "demo": false, "description": "Claw-Anything evaluates always-on LLM personal assistants in simulated environments with long-horizon activity histories, interdependent backend services, and GUI/CLI across devices. It includes 200 human-verified tasks scored on completion, robustness, communication, and safety, with a live leaderboard.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence"], "snippet": "To address this gap, we introduce Claw-Anything, a benchmark that expands agent context along three dimensions: long-horizon activity histories, interdependent backend services, and integrated GUI and CLI interaction across multiple devices."}, "familyId": "bmf_ebd37017ecf1", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 1, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [{"evidence": [{"basis": "existing-code-link", "linkedUrl": "https://github.com/LiberCoders/Claw-Anything", "url": "https://github.com/LiberCoders/Claw-Anything"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/LiberCoders/Claw-Anything", "url": "https://huggingface.co/papers/2605.26086"}, {"basis": "huggingface-paper-association", "linkedUrl": "https://github.com/LiberCoders/CLaw-Anything", "url": "https://huggingface.co/api/papers/2605.26086"}, {"basis": "linked-from-project", "linkedUrl": "https://github.com/LiberCoders/Claw-Anything", "url": "https://libercoders.github.io/Claw-Anything/"}, {"basis": "linked-from-data", "linkedUrl": "https://github.com/LiberCoders/CLaw-Anything", "url": "https://huggingface.co/datasets/LiberCoders/Claw-Anything"}, {"basis": "linked-from-data", "linkedUrl": "https://github.com/LiberCoders/CLaw-Anything.git", "url": "https://huggingface.co/datasets/LiberCoders/Claw-Anything"}], "scope": "benchmark_repo", "url": "https://github.com/LiberCoders/Claw-Anything", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_claw-anything_3c56b6b4", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/LiberCoders/Claw-Anything", "data": "https://huggingface.co/datasets/LiberCoders/Claw-Anything", "hfPaper": "https://huggingface.co/papers/2605.26086", "paper": "https://arxiv.org/abs/2605.26086", "pdf": "https://arxiv.org/pdf/2605.26086", "project": "https://libercoders.github.io/Claw-Anything/", "report": "https://arxiv.org/abs/2605.26086"}, "metrics": [], "motivation": "Large language model agents are increasingly envisioned as always-on personal assistants with access to anything relevant in the user's digital world.", "name": "Claw-Anything", "oneLine": "Claw-Anything evaluates always-on LLM personal assistants in simulated environments with long-horizon activity histories, interdependent backend services, and GUI/CLI across devices. It includes 200 human-verified tasks scored on completion, robustness, communication, and safety, with a live leaderboard.", "paperTitle": "Claw-Anything: Benchmarking Always-On Personal Assistants with Broader Access to User's Digital World", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2605.26086", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "LiberCoders", "organizationType": "community", "role": "benchmark-publisher", "sourceUrl": "https://github.com/LiberCoders/Claw-Anything"}], "ranking": {}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releaseDates": {"firstPublicAt": "2026-05-25", "paperV1At": "2026-05-25"}, "releasedAt": "2026-05-25", "source": {"authors": ["Yusong Lin", "Xinyuan Liang", "Haiyang Wang", "Qipeng Gu", "Siqi Cheng", "Jiangui Chen", "Shuzhe Wu", "Feiyang Pan", "Lue Fan", "Sanyuan Zhao", "Dandan Tu"], "categories": ["cs.AI"], "id": "2605.26086", "title": "Claw-Anything: Benchmarking Always-On Personal Assistants with Broader Access to User's Digital World", "type": "arxiv", "url": "https://arxiv.org/abs/2605.26086"}, "sourceUpdatedAt": "2026-05-25T17:50:04Z", "topics": ["Agents"], "venueAttempts": [], "whyItMatters": "It expands agent evaluation to broad, always-on contexts, revealing capability gaps in stateful, proactive assistance and supporting scalable data generation for training."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
