{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UN6S3554337PHGRE37RU4QTKV6","short_pith_number":"pith:UN6S3554","schema_version":"1.0","canonical_sha256":"a37d2df7bcdefef39a24dfe34e426aaf86375ccd1f8335b6ec1209cdcd0aefe0","source":{"kind":"arxiv","id":"2606.05233","version":1},"attestation_state":"computed","paper":{"title":"Domain-Conditioned Safety in Frontier Computer-Using Agents: A 793-Episode Browser Benchmark, a Coding-Domain Cross-Reference, and a Reproducibility Audit of Recent Red-Teaming","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CR","authors_text":"Nicholas Saban","submitted_at":"2026-06-03T01:21:59Z","abstract_excerpt":"Recent computer-using-agent (CUA) red-teaming papers report prompt-injection attack success rates (ASR) of 42-98%, but these headline numbers cluster on retired models and on the most-vulnerable model in each paper's panel. We ask whether those techniques, reproduced as hand-crafted templates, still work against current frontier CUAs. We release CUA-HandCrafted, a public benchmark of 793 episodes spanning 24 multi-step web tasks, 56 attack templates, 8 attack families, and 4 system-prompt configurations. Against Claude Sonnet 4.6 and GPT-5.4 we measure 0/140 multi-step attack success (Clopper-"},"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":"2606.05233","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-06-03T01:21:59Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"eb7caf708bf4a32cac154ef9ae25d41ca812cff1ce8be4a4e020e270179c8d70","abstract_canon_sha256":"562be3f450e747eb225948c34fce90135d5a4cde19512c60fd91f4dbe64e7cb0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T00:13:49.554713Z","signature_b64":"daTMAEF4XLDvOXcJoO/sLZm/WKLoVPFuAXqIaIuvXSA81Ha+iuIHkp0O1MM4dKNfhklDY2rRjorhAwFmSAq+BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a37d2df7bcdefef39a24dfe34e426aaf86375ccd1f8335b6ec1209cdcd0aefe0","last_reissued_at":"2026-06-05T00:13:49.554225Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T00:13:49.554225Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Domain-Conditioned Safety in Frontier Computer-Using Agents: A 793-Episode Browser Benchmark, a Coding-Domain Cross-Reference, and a Reproducibility Audit of Recent Red-Teaming","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CR","authors_text":"Nicholas Saban","submitted_at":"2026-06-03T01:21:59Z","abstract_excerpt":"Recent computer-using-agent (CUA) red-teaming papers report prompt-injection attack success rates (ASR) of 42-98%, but these headline numbers cluster on retired models and on the most-vulnerable model in each paper's panel. We ask whether those techniques, reproduced as hand-crafted templates, still work against current frontier CUAs. We release CUA-HandCrafted, a public benchmark of 793 episodes spanning 24 multi-step web tasks, 56 attack templates, 8 attack families, and 4 system-prompt configurations. Against Claude Sonnet 4.6 and GPT-5.4 we measure 0/140 multi-step attack success (Clopper-"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.05233","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/2606.05233/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":"2606.05233","created_at":"2026-06-05T00:13:49.554295+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.05233v1","created_at":"2026-06-05T00:13:49.554295+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.05233","created_at":"2026-06-05T00:13:49.554295+00:00"},{"alias_kind":"pith_short_12","alias_value":"UN6S3554337P","created_at":"2026-06-05T00:13:49.554295+00:00"},{"alias_kind":"pith_short_16","alias_value":"UN6S3554337PHGRE","created_at":"2026-06-05T00:13:49.554295+00:00"},{"alias_kind":"pith_short_8","alias_value":"UN6S3554","created_at":"2026-06-05T00:13:49.554295+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/UN6S3554337PHGRE37RU4QTKV6","json":"https://pith.science/pith/UN6S3554337PHGRE37RU4QTKV6.json","graph_json":"https://pith.science/api/pith-number/UN6S3554337PHGRE37RU4QTKV6/graph.json","events_json":"https://pith.science/api/pith-number/UN6S3554337PHGRE37RU4QTKV6/events.json","paper":"https://pith.science/paper/UN6S3554"},"agent_actions":{"view_html":"https://pith.science/pith/UN6S3554337PHGRE37RU4QTKV6","download_json":"https://pith.science/pith/UN6S3554337PHGRE37RU4QTKV6.json","view_paper":"https://pith.science/paper/UN6S3554","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.05233&json=true","fetch_graph":"https://pith.science/api/pith-number/UN6S3554337PHGRE37RU4QTKV6/graph.json","fetch_events":"https://pith.science/api/pith-number/UN6S3554337PHGRE37RU4QTKV6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UN6S3554337PHGRE37RU4QTKV6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UN6S3554337PHGRE37RU4QTKV6/action/storage_attestation","attest_author":"https://pith.science/pith/UN6S3554337PHGRE37RU4QTKV6/action/author_attestation","sign_citation":"https://pith.science/pith/UN6S3554337PHGRE37RU4QTKV6/action/citation_signature","submit_replication":"https://pith.science/pith/UN6S3554337PHGRE37RU4QTKV6/action/replication_record"}},"created_at":"2026-06-05T00:13:49.554295+00:00","updated_at":"2026-06-05T00:13:49.554295+00:00"}