{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:FLB7TK54JF4GASY34REVHEF5LN","short_pith_number":"pith:FLB7TK54","schema_version":"1.0","canonical_sha256":"2ac3f9abbc4978604b1be4495390bd5b450999a5cb30708d9eab014d398aa30c","source":{"kind":"arxiv","id":"2606.32007","version":1},"attestation_state":"computed","paper":{"title":"AxDafny: Agentic Verified Code Generation in Dafny","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Austin Letson, Benjamin Breen, Borja Requena Pozo, Leopoldo Sarra","submitted_at":"2026-06-30T17:39:43Z","abstract_excerpt":"We study agentic code generation in Dafny, where a model must generate both executable code and the proof artifacts for verification. We present AxDafny, a verifier-guided repair framework that iteratively generates implementations, invariants, assertions, and termination arguments. We also introduce LiveCodeBench-Pro-Dafny (LCB-Pro-Dafny), a benchmark of 250 competition-style programming problems translated into Dafny with formal specifications and a verifier-based evaluation harness. On LCB-Pro-Dafny, AxDafny substantially improves verification success over baseline GPT-5.5 performance. On D"},"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":"2606.32007","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-30T17:39:43Z","cross_cats_sorted":[],"title_canon_sha256":"6293764529dba101066bd5d1a03ae3523d3c6460ff52ff36c0473d3fa35bd448","abstract_canon_sha256":"8e4b3d93881ba00912b03fbdf56cc1298d3755ce63bcfa6e12da5cae7e984067"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-01T01:18:28.898790Z","signature_b64":"XCOLBujWK2g5px2dvoaF7mZyiOqhdotWzUjcnUMdla3RyZstjXbsmXtnTuGIWV9GuFx7QrxdxzJm4sQ9KYJnBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2ac3f9abbc4978604b1be4495390bd5b450999a5cb30708d9eab014d398aa30c","last_reissued_at":"2026-07-01T01:18:28.898382Z","signature_status":"signed_v1","first_computed_at":"2026-07-01T01:18:28.898382Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AxDafny: Agentic Verified Code Generation in Dafny","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Austin Letson, Benjamin Breen, Borja Requena Pozo, Leopoldo Sarra","submitted_at":"2026-06-30T17:39:43Z","abstract_excerpt":"We study agentic code generation in Dafny, where a model must generate both executable code and the proof artifacts for verification. We present AxDafny, a verifier-guided repair framework that iteratively generates implementations, invariants, assertions, and termination arguments. We also introduce LiveCodeBench-Pro-Dafny (LCB-Pro-Dafny), a benchmark of 250 competition-style programming problems translated into Dafny with formal specifications and a verifier-based evaluation harness. On LCB-Pro-Dafny, AxDafny substantially improves verification success over baseline GPT-5.5 performance. On D"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.32007","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/2606.32007/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":"2606.32007","created_at":"2026-07-01T01:18:28.898443+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.32007v1","created_at":"2026-07-01T01:18:28.898443+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.32007","created_at":"2026-07-01T01:18:28.898443+00:00"},{"alias_kind":"pith_short_12","alias_value":"FLB7TK54JF4G","created_at":"2026-07-01T01:18:28.898443+00:00"},{"alias_kind":"pith_short_16","alias_value":"FLB7TK54JF4GASY3","created_at":"2026-07-01T01:18:28.898443+00:00"},{"alias_kind":"pith_short_8","alias_value":"FLB7TK54","created_at":"2026-07-01T01:18:28.898443+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/FLB7TK54JF4GASY34REVHEF5LN","json":"https://pith.science/pith/FLB7TK54JF4GASY34REVHEF5LN.json","graph_json":"https://pith.science/api/pith-number/FLB7TK54JF4GASY34REVHEF5LN/graph.json","events_json":"https://pith.science/api/pith-number/FLB7TK54JF4GASY34REVHEF5LN/events.json","paper":"https://pith.science/paper/FLB7TK54"},"agent_actions":{"view_html":"https://pith.science/pith/FLB7TK54JF4GASY34REVHEF5LN","download_json":"https://pith.science/pith/FLB7TK54JF4GASY34REVHEF5LN.json","view_paper":"https://pith.science/paper/FLB7TK54","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.32007&json=true","fetch_graph":"https://pith.science/api/pith-number/FLB7TK54JF4GASY34REVHEF5LN/graph.json","fetch_events":"https://pith.science/api/pith-number/FLB7TK54JF4GASY34REVHEF5LN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FLB7TK54JF4GASY34REVHEF5LN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FLB7TK54JF4GASY34REVHEF5LN/action/storage_attestation","attest_author":"https://pith.science/pith/FLB7TK54JF4GASY34REVHEF5LN/action/author_attestation","sign_citation":"https://pith.science/pith/FLB7TK54JF4GASY34REVHEF5LN/action/citation_signature","submit_replication":"https://pith.science/pith/FLB7TK54JF4GASY34REVHEF5LN/action/replication_record"}},"created_at":"2026-07-01T01:18:28.898443+00:00","updated_at":"2026-07-01T01:18:28.898443+00:00"}