{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:SZBYRJKINWEULNIWRPPUFF33OT","short_pith_number":"pith:SZBYRJKI","canonical_record":{"source":{"id":"2607.20911","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-23T04:34:06Z","cross_cats_sorted":["cs.SE"],"title_canon_sha256":"e53c8a2c08e59c65e79458bf51c68a35f0e0154bbba2db6e211b3d463e4fe92b","abstract_canon_sha256":"5b1d8b9e04bd4d828b013b43cd5bc564a88efbc52d19b21ab1820e4e3a5cc3a1"},"schema_version":"1.0"},"canonical_sha256":"964388a5486d8945b5168bdf42977b74d73f66593398239d4d30e2a1028725ce","source":{"kind":"arxiv","id":"2607.20911","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.20911","created_at":"2026-07-24T01:23:40Z"},{"alias_kind":"arxiv_version","alias_value":"2607.20911v1","created_at":"2026-07-24T01:23:40Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.20911","created_at":"2026-07-24T01:23:40Z"},{"alias_kind":"pith_short_12","alias_value":"SZBYRJKINWEU","created_at":"2026-07-24T01:23:40Z"},{"alias_kind":"pith_short_16","alias_value":"SZBYRJKINWEULNIW","created_at":"2026-07-24T01:23:40Z"},{"alias_kind":"pith_short_8","alias_value":"SZBYRJKI","created_at":"2026-07-24T01:23:40Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:SZBYRJKINWEULNIWRPPUFF33OT","target":"record","payload":{"canonical_record":{"source":{"id":"2607.20911","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-23T04:34:06Z","cross_cats_sorted":["cs.SE"],"title_canon_sha256":"e53c8a2c08e59c65e79458bf51c68a35f0e0154bbba2db6e211b3d463e4fe92b","abstract_canon_sha256":"5b1d8b9e04bd4d828b013b43cd5bc564a88efbc52d19b21ab1820e4e3a5cc3a1"},"schema_version":"1.0"},"canonical_sha256":"964388a5486d8945b5168bdf42977b74d73f66593398239d4d30e2a1028725ce","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-24T01:23:40.350253Z","signature_b64":"jyuUAS2zBgpxsB+2dpgUyurEIYlWB33gGfosMssTYdRGhHfkMAGzLKLzty2njclH0gpZKRi/MTG8hN/wrXpqBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"964388a5486d8945b5168bdf42977b74d73f66593398239d4d30e2a1028725ce","last_reissued_at":"2026-07-24T01:23:40.349285Z","signature_status":"signed_v1","first_computed_at":"2026-07-24T01:23:40.349285Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.20911","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-24T01:23:40Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5AcZvsAyImwOhyFf68oHZM5YjWp/lDtR+cyF7kgO5QTWp5kVbgS3hu630JH5CdyHV9ktUBN9eM6cshXoSatNBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T23:40:59.958324Z"},"content_sha256":"86954c1a56dacbe89ca2fd3acbe255ec7e6be9066590ea6af8559ae67e192880","schema_version":"1.0","event_id":"sha256:86954c1a56dacbe89ca2fd3acbe255ec7e6be9066590ea6af8559ae67e192880"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:SZBYRJKINWEULNIWRPPUFF33OT","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Tencent WorkBuddy Bench: A Multi-Domain Coding-Agent Benchmark with Contamination-Resistant Task Construction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.SE"],"primary_cat":"cs.CL","authors_text":"Anyuan Sang, Chaofan Qiu, Chaofan Zhu, Feng Wu, Hande Dong, Kai Zhang, Ke Li, Ling Jiang, Linjie Che, Ning Yang, Peijie Dong, Qiang Lin, Sen Nie, Shaopeng Chen, Shi Wu, Shuwen Zhang, Tencent WorkBuddy Bench Team: Siqi Cai, Xiang Fei, Xianjun Zhang, Xiaoxi Zhao, Xiaoyun Liang, Xing Sun, Xin Li, Yi Liu, Yong Mao, Yuanjie Zhu, Yuanzhen Xu, Yubin Qi, Yuchen Shi, Yunhao Zhang, Yunsheng Wu, Zhehang Chu, Zhen Feng, Zhiheng Lyu, Zhijian Shao, Zhiqiang Dong, Zihan Xu","submitted_at":"2026-07-23T04:34:06Z","abstract_excerpt":"We introduce Tencent WorkBuddy Bench, a multi-domain evaluation suite for coding agents; this report documents its construction methodology, scoring protocol, and a cross-model leaderboard. At its core is a unified evaluation framework for constructing and running distribution-informed coding-agent tasks across four work domains - Code, Web, Office, and Security. Rather than adapting public issue text, every task is reverse-engineered from a real commit, pull request, or business scenario and rewritten as a short, colloquial, role-played request, so that a task's prompt is not recoverable by w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.20911","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/2607.20911/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-24T01:23:40Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RhdeAP7heyom+BwmGsGJhfYwXSz816Shx4ImnjHRDSGQRgUb3kiOu3uNoyWZfy5GfMXBiRzC2hJjtDu8Z2DzAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T23:40:59.959256Z"},"content_sha256":"3cafb5aca67eaa9a315cc8806e3459e0d0c3ec8af20118a3632670fe67938304","schema_version":"1.0","event_id":"sha256:3cafb5aca67eaa9a315cc8806e3459e0d0c3ec8af20118a3632670fe67938304"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/SZBYRJKINWEULNIWRPPUFF33OT/bundle.json","state_url":"https://pith.science/pith/SZBYRJKINWEULNIWRPPUFF33OT/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/SZBYRJKINWEULNIWRPPUFF33OT/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-17T23:40:59Z","links":{"resolver":"https://pith.science/pith/SZBYRJKINWEULNIWRPPUFF33OT","bundle":"https://pith.science/pith/SZBYRJKINWEULNIWRPPUFF33OT/bundle.json","state":"https://pith.science/pith/SZBYRJKINWEULNIWRPPUFF33OT/state.json","well_known_bundle":"https://pith.science/.well-known/pith/SZBYRJKINWEULNIWRPPUFF33OT/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:SZBYRJKINWEULNIWRPPUFF33OT","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"5b1d8b9e04bd4d828b013b43cd5bc564a88efbc52d19b21ab1820e4e3a5cc3a1","cross_cats_sorted":["cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-23T04:34:06Z","title_canon_sha256":"e53c8a2c08e59c65e79458bf51c68a35f0e0154bbba2db6e211b3d463e4fe92b"},"schema_version":"1.0","source":{"id":"2607.20911","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.20911","created_at":"2026-07-24T01:23:40Z"},{"alias_kind":"arxiv_version","alias_value":"2607.20911v1","created_at":"2026-07-24T01:23:40Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.20911","created_at":"2026-07-24T01:23:40Z"},{"alias_kind":"pith_short_12","alias_value":"SZBYRJKINWEU","created_at":"2026-07-24T01:23:40Z"},{"alias_kind":"pith_short_16","alias_value":"SZBYRJKINWEULNIW","created_at":"2026-07-24T01:23:40Z"},{"alias_kind":"pith_short_8","alias_value":"SZBYRJKI","created_at":"2026-07-24T01:23:40Z"}],"graph_snapshots":[{"event_id":"sha256:3cafb5aca67eaa9a315cc8806e3459e0d0c3ec8af20118a3632670fe67938304","target":"graph","created_at":"2026-07-24T01:23:40Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2607.20911/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We introduce Tencent WorkBuddy Bench, a multi-domain evaluation suite for coding agents; this report documents its construction methodology, scoring protocol, and a cross-model leaderboard. At its core is a unified evaluation framework for constructing and running distribution-informed coding-agent tasks across four work domains - Code, Web, Office, and Security. Rather than adapting public issue text, every task is reverse-engineered from a real commit, pull request, or business scenario and rewritten as a short, colloquial, role-played request, so that a task's prompt is not recoverable by w","authors_text":"Anyuan Sang, Chaofan Qiu, Chaofan Zhu, Feng Wu, Hande Dong, Kai Zhang, Ke Li, Ling Jiang, Linjie Che, Ning Yang, Peijie Dong, Qiang Lin, Sen Nie, Shaopeng Chen, Shi Wu, Shuwen Zhang, Tencent WorkBuddy Bench Team: Siqi Cai, Xiang Fei, Xianjun Zhang, Xiaoxi Zhao, Xiaoyun Liang, Xing Sun, Xin Li, Yi Liu, Yong Mao, Yuanjie Zhu, Yuanzhen Xu, Yubin Qi, Yuchen Shi, Yunhao Zhang, Yunsheng Wu, Zhehang Chu, Zhen Feng, Zhiheng Lyu, Zhijian Shao, Zhiqiang Dong, Zihan Xu","cross_cats":["cs.SE"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-23T04:34:06Z","title":"Tencent WorkBuddy Bench: A Multi-Domain Coding-Agent Benchmark with Contamination-Resistant Task Construction"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.20911","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:86954c1a56dacbe89ca2fd3acbe255ec7e6be9066590ea6af8559ae67e192880","target":"record","created_at":"2026-07-24T01:23:40Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"5b1d8b9e04bd4d828b013b43cd5bc564a88efbc52d19b21ab1820e4e3a5cc3a1","cross_cats_sorted":["cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-23T04:34:06Z","title_canon_sha256":"e53c8a2c08e59c65e79458bf51c68a35f0e0154bbba2db6e211b3d463e4fe92b"},"schema_version":"1.0","source":{"id":"2607.20911","kind":"arxiv","version":1}},"canonical_sha256":"964388a5486d8945b5168bdf42977b74d73f66593398239d4d30e2a1028725ce","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"964388a5486d8945b5168bdf42977b74d73f66593398239d4d30e2a1028725ce","first_computed_at":"2026-07-24T01:23:40.349285Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-24T01:23:40.349285Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"jyuUAS2zBgpxsB+2dpgUyurEIYlWB33gGfosMssTYdRGhHfkMAGzLKLzty2njclH0gpZKRi/MTG8hN/wrXpqBg==","signature_status":"signed_v1","signed_at":"2026-07-24T01:23:40.350253Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.20911","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:86954c1a56dacbe89ca2fd3acbe255ec7e6be9066590ea6af8559ae67e192880","sha256:3cafb5aca67eaa9a315cc8806e3459e0d0c3ec8af20118a3632670fe67938304"],"state_sha256":"d4dacde880de265d4d747239728bf9c912ceb45d4203a2298a18695a654dc9ab"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"P+bx9+FzOSxlbD9IdDDFQ6Pc844PK9Bkck2I1jPdgXlv4psgyVumsw8q54/W+clH1fyr6uWZI1wKRympDZRNDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-17T23:40:59.965009Z","bundle_sha256":"6993f2852fa772bbbee08d81648b15bcdce6ef0f89dd481deb42a1f784ee1031"}}