{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:PPJGS2D3HRN5L2LXZXRSOKIZ7M","short_pith_number":"pith:PPJGS2D3","schema_version":"1.0","canonical_sha256":"7bd269687b3c5bd5e977cde3272919fb2cb22ac215ecfea37495cb68b490f7f2","source":{"kind":"arxiv","id":"2608.09254","version":1},"attestation_state":"computed","paper":{"title":"Business Truth, not SQL Accuracy: A Rule-Gated 7B Analytics Agent Outperforms a Direct-Prompted 32B Baseline","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Morris Lee","submitted_at":"2026-08-10T08:14:24Z","abstract_excerpt":"LLM analytics agents are evaluated on SQL syntax accuracy, but production failures look different: questions with two valid business definitions, questions the warehouse cannot answer, deprecated columns after a schema change, and queries that execute successfully while returning the wrong business number. No execution-match metric can score them. This paper introduces WarehouseReliabilityBench, 400 frozen tasks over two synthetic warehouses in which roughly half the correct responses are a clarification, an abstention or a refusal, with pinned denominators and a pre-registered paired bootstra"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2608.09254","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-10T08:14:24Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"1f43d154f6b055e0361ba77f63f078b0efbc8fb10b4ea70d71874be3a192d77e","abstract_canon_sha256":"6e5b04d502f6061224d4cb37493344abdb44a037e5ab7152deb57fc49b1c5cb5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T02:22:11.268039Z","signature_b64":"7awgbuSwaNemtbmo7m+U93YVI7bsNpC4wqWnoXzbGqwp0Tq8LHISi2Xp9FiRTt/XFr85ucdLbT+QrEkw58+pCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7bd269687b3c5bd5e977cde3272919fb2cb22ac215ecfea37495cb68b490f7f2","last_reissued_at":"2026-08-11T02:22:11.266585Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T02:22:11.266585Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Business Truth, not SQL Accuracy: A Rule-Gated 7B Analytics Agent Outperforms a Direct-Prompted 32B Baseline","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Morris Lee","submitted_at":"2026-08-10T08:14:24Z","abstract_excerpt":"LLM analytics agents are evaluated on SQL syntax accuracy, but production failures look different: questions with two valid business definitions, questions the warehouse cannot answer, deprecated columns after a schema change, and queries that execute successfully while returning the wrong business number. No execution-match metric can score them. This paper introduces WarehouseReliabilityBench, 400 frozen tasks over two synthetic warehouses in which roughly half the correct responses are a clarification, an abstention or a refusal, with pinned denominators and a pre-registered paired bootstra"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.09254","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2608.09254/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2608.09254","created_at":"2026-08-11T02:22:11.267165+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.09254v1","created_at":"2026-08-11T02:22:11.267165+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.09254","created_at":"2026-08-11T02:22:11.267165+00:00"},{"alias_kind":"pith_short_12","alias_value":"PPJGS2D3HRN5","created_at":"2026-08-11T02:22:11.267165+00:00"},{"alias_kind":"pith_short_16","alias_value":"PPJGS2D3HRN5L2LX","created_at":"2026-08-11T02:22:11.267165+00:00"},{"alias_kind":"pith_short_8","alias_value":"PPJGS2D3","created_at":"2026-08-11T02:22:11.267165+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PPJGS2D3HRN5L2LXZXRSOKIZ7M","json":"https://pith.science/pith/PPJGS2D3HRN5L2LXZXRSOKIZ7M.json","graph_json":"https://pith.science/api/pith-number/PPJGS2D3HRN5L2LXZXRSOKIZ7M/graph.json","events_json":"https://pith.science/api/pith-number/PPJGS2D3HRN5L2LXZXRSOKIZ7M/events.json","paper":"https://pith.science/paper/PPJGS2D3"},"agent_actions":{"view_html":"https://pith.science/pith/PPJGS2D3HRN5L2LXZXRSOKIZ7M","download_json":"https://pith.science/pith/PPJGS2D3HRN5L2LXZXRSOKIZ7M.json","view_paper":"https://pith.science/paper/PPJGS2D3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.09254&json=true","fetch_graph":"https://pith.science/api/pith-number/PPJGS2D3HRN5L2LXZXRSOKIZ7M/graph.json","fetch_events":"https://pith.science/api/pith-number/PPJGS2D3HRN5L2LXZXRSOKIZ7M/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PPJGS2D3HRN5L2LXZXRSOKIZ7M/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PPJGS2D3HRN5L2LXZXRSOKIZ7M/action/storage_attestation","attest_author":"https://pith.science/pith/PPJGS2D3HRN5L2LXZXRSOKIZ7M/action/author_attestation","sign_citation":"https://pith.science/pith/PPJGS2D3HRN5L2LXZXRSOKIZ7M/action/citation_signature","submit_replication":"https://pith.science/pith/PPJGS2D3HRN5L2LXZXRSOKIZ7M/action/replication_record"}},"created_at":"2026-08-11T02:22:11.267165+00:00","updated_at":"2026-08-11T02:22:11.267165+00:00"}