{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2608.29128", "kind": "paper", "url": "https://arxiv.org/abs/2608.29128"}, {"id": "gh:postmanlabs/apiflow-bench", "kind": "repo", "url": "https://github.com/postmanlabs/APIFlow-Bench"}], "authors": ["Zelin Wan", "Arash Nourian", "Xiaoxiao Li", "Nihar Nandan", "Kamalakannan Nandagopal"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.29128", "source_url": "https://arxiv.org/abs/2608.29128"}, "categories": ["General AI", "Language & Knowledge", "cs.AI"], "description": {"en": "Evaluates tool-using agents on 467 long-horizon REST API tasks with deterministic provenance-sensitive grading, measuring reliability, efficiency, robustness, and other engineering capabilities."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.29128", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.29128", "title": "APIFlow-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2608.29128", "modality": null, "name": "APIFlow-Bench", "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": "true", "origin_record_id": "bm_apiflow-bench_b256ba9f", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "e815696933ad9eaa2f4863a984d830f8fb306eec808c03a852c706d8fc9b24e8", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-02T05:23:11.715633Z", "source_url": "https://arxiv.org/abs/2608.29128"}, "publisher": null, "released": "2026-08-29", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.29128", "source_url": "https://arxiv.org/abs/2608.29128"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-29128", "source": "claire_radar", "source_benchmark_id": "2608.29128", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Machine Generated", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-09-02", "attemptedAt": "2026-09-02T05:12:51.621571Z", "githubRepo": "https://github.com/postmanlabs/APIFlow-Bench", "githubScope": "benchmark_repo", "githubStars": 4, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2608.29128", "observedAt": "2026-09-02T05:12:51.621571Z", "signalStatus": {"githubStars": {"observedAt": "2026-09-02T05:12:51.621571Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-02T05:12:51.621571Z", "state": "unavailable"}}, "status": "fresh"}, "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The benchmark targets a highly active area of tool-using agents and offers complete public infrastructure including transcripts and leaderboards, which strongly drives immediate community engagement.", "score": 82}, "capabilities": [], "confidence": "High", "construction": "Original Synthetic", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-02T05:23:11.715633Z", "inputHash": "a489a5283bb8d4b0c838455eb23d64fb47acfd713645a55a67922026de50c2bf", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "We introduce APIFlow-Bench, a fully auditable benchmark for long-horizon, dependent REST-API workflows that decomposes performance into seven engineering capabilities and requires agents to produce answers supported by the actual call path.", "canonicalNameSource": "abstract", "decisionReason": "Formally named benchmark with a frozen task bank, detailed scoring, public code, leaderboard, and extensive release artifacts.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-02T05:23:11.715633Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates tool-using agents on 467 long-horizon REST API tasks with deterministic provenance-sensitive grading, measuring reliability, efficiency, robustness, and other engineering capabilities.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "score_submission", "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL"], "snippet": "We introduce APIFlow-Bench, a fully auditable benchmark for long-horizon, dependent REST-API workflows that decomposes performance into seven engineering capabilities and requires agents to produce answers supported by the actual call path."}, "familyId": "bmf_089c7bca50fb", "firstSeenAt": "2026-09-01", "heat": null, "id": "bm_apiflow-bench_b256ba9f", "indexedAt": "2026-09-01T06:28:29.895112Z", "industrySectors": [], "links": {"code": "https://github.com/postmanlabs/APIFlow-Bench", "data": null, "hfPaper": "https://huggingface.co/papers/2608.29128", "paper": "https://arxiv.org/abs/2608.29128", "pdf": "https://arxiv.org/pdf/2608.29128", "project": null, "report": "https://arxiv.org/abs/2608.29128"}, "metrics": [], "motivation": "Tool-using agents are commonly evaluated by a single bit: whether an end-to-end workflow completed.", "name": "APIFlow-Bench", "oneLine": "Evaluates tool-using agents on 467 long-horizon REST API tasks with deterministic provenance-sensitive grading, measuring reliability, efficiency, robustness, and other engineering capabilities.", "paperTitle": "APIFlow-Bench: Measuring Whether Agents Survive Long, Dependent API Workflows", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2608.29128", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-02T00:50:41.903960Z"}, "publications": [], "publishers": [{"name": "Postman Labs", "organizationType": "company-research-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/postmanlabs/APIFlow-Bench"}], "ranking": {"30d": {"confidence": "Low", "coverage": 0.55, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 30}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": 0.2397076310863079, "value": 4}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 65, "score": 36}, "method": "attention-ranking-v13", "rank": 65, "score": 36}, "90d": {"confidence": "Low", "coverage": 0.55, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 90}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": 0.21905162141015003, "value": 4}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 194, "score": 35}, "method": "attention-ranking-v13", "rank": 194, "score": 35}}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-08-29", "source": {"authors": ["Zelin Wan", "Arash Nourian", "Xiaoxiao Li", "Nihar Nandan", "Kamalakannan Nandagopal"], "categories": ["cs.AI", "cs.LG", "cs.SE"], "id": "2608.29128", "title": "APIFlow-Bench: Measuring Whether Agents Survive Long, Dependent API Workflows", "type": "arxiv", "url": "https://arxiv.org/abs/2608.29128"}, "sourceUpdatedAt": "2026-08-29T08:12:11Z", "topics": ["cs.AI"], "venueAttempts": [], "whyItMatters": "Provides a production-oriented benchmark that distinguishes meaningful failure modes and offers full transparency with public transcripts and a leaderboard, enabling robust agent comparisons."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
