{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ZNN56TI3HFB4OZ3JINVHATCUET","short_pith_number":"pith:ZNN56TI3","schema_version":"1.0","canonical_sha256":"cb5bdf4d1b3943c76769436a704c5424fc816e865156f08a741a9e9004835bf9","source":{"kind":"arxiv","id":"2608.09567","version":1},"attestation_state":"computed","paper":{"title":"From Runnable to Verifiable: An Independent Reproducibility Study of LLM/Agent-Driven Vulnerability Validation Artifacts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CR","authors_text":"Bo Chen","submitted_at":"2026-08-10T13:05:02Z","abstract_excerpt":"Security research artifacts---repositories, PoC exploits, and validation pipelines---are increasingly produced by LLM/agent-driven vulnerability workflows, yet the gap between \\emph{publicly available}, \\emph{runnable}, \\emph{signal-producing}, and \\emph{semantically confirmed} artifacts is poorly measured. We conduct a pre-registered reproducibility audit of this literature. A search covering 2023--2026 with dual screening yields a 104-paper consensus corpus, of which 59 papers (56.7\\%) have a publicly reachable artifact. We execute an 18-paper sample at R0/R1 and all 102 cases of the anchor "},"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":"2608.09567","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-08-10T13:05:02Z","cross_cats_sorted":[],"title_canon_sha256":"f33c6abac95b4c919d1ea72362abc31693d3d04eb380d5315c7548f917df7639","abstract_canon_sha256":"973f285f83a7a13c19f1e01a0ea18011798b0c27051ee306b1d0528feab2f043"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T02:24:31.122082Z","signature_b64":"RvjmaEFWTj4Rozt1JCDbZHBKk1lF3BVyX2RiVXZMRaLu/E4tth68+fwDy8GfKaT2McSx9cNqz5YxVH4Htk/0Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cb5bdf4d1b3943c76769436a704c5424fc816e865156f08a741a9e9004835bf9","last_reissued_at":"2026-08-11T02:24:31.119982Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T02:24:31.119982Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From Runnable to Verifiable: An Independent Reproducibility Study of LLM/Agent-Driven Vulnerability Validation Artifacts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CR","authors_text":"Bo Chen","submitted_at":"2026-08-10T13:05:02Z","abstract_excerpt":"Security research artifacts---repositories, PoC exploits, and validation pipelines---are increasingly produced by LLM/agent-driven vulnerability workflows, yet the gap between \\emph{publicly available}, \\emph{runnable}, \\emph{signal-producing}, and \\emph{semantically confirmed} artifacts is poorly measured. We conduct a pre-registered reproducibility audit of this literature. A search covering 2023--2026 with dual screening yields a 104-paper consensus corpus, of which 59 papers (56.7\\%) have a publicly reachable artifact. We execute an 18-paper sample at R0/R1 and all 102 cases of the anchor "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.09567","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/2608.09567/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":"2608.09567","created_at":"2026-08-11T02:24:31.120831+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.09567v1","created_at":"2026-08-11T02:24:31.120831+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.09567","created_at":"2026-08-11T02:24:31.120831+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZNN56TI3HFB4","created_at":"2026-08-11T02:24:31.120831+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZNN56TI3HFB4OZ3J","created_at":"2026-08-11T02:24:31.120831+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZNN56TI3","created_at":"2026-08-11T02:24:31.120831+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/ZNN56TI3HFB4OZ3JINVHATCUET","json":"https://pith.science/pith/ZNN56TI3HFB4OZ3JINVHATCUET.json","graph_json":"https://pith.science/api/pith-number/ZNN56TI3HFB4OZ3JINVHATCUET/graph.json","events_json":"https://pith.science/api/pith-number/ZNN56TI3HFB4OZ3JINVHATCUET/events.json","paper":"https://pith.science/paper/ZNN56TI3"},"agent_actions":{"view_html":"https://pith.science/pith/ZNN56TI3HFB4OZ3JINVHATCUET","download_json":"https://pith.science/pith/ZNN56TI3HFB4OZ3JINVHATCUET.json","view_paper":"https://pith.science/paper/ZNN56TI3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.09567&json=true","fetch_graph":"https://pith.science/api/pith-number/ZNN56TI3HFB4OZ3JINVHATCUET/graph.json","fetch_events":"https://pith.science/api/pith-number/ZNN56TI3HFB4OZ3JINVHATCUET/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZNN56TI3HFB4OZ3JINVHATCUET/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZNN56TI3HFB4OZ3JINVHATCUET/action/storage_attestation","attest_author":"https://pith.science/pith/ZNN56TI3HFB4OZ3JINVHATCUET/action/author_attestation","sign_citation":"https://pith.science/pith/ZNN56TI3HFB4OZ3JINVHATCUET/action/citation_signature","submit_replication":"https://pith.science/pith/ZNN56TI3HFB4OZ3JINVHATCUET/action/replication_record"}},"created_at":"2026-08-11T02:24:31.120831+00:00","updated_at":"2026-08-11T02:24:31.120831+00:00"}