{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:K32DAGUIUKS7XYYOPJW2IMTKV2","short_pith_number":"pith:K32DAGUI","schema_version":"1.0","canonical_sha256":"56f4301a88a2a5fbe30e7a6da4326aae800c853c1a3ec9f7854c7e4d1fc1a31e","source":{"kind":"arxiv","id":"2607.06184","version":1},"attestation_state":"computed","paper":{"title":"What Resolve Rate Hides: Trajectory Structure Diagnostics for Coding Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Chun Yong Chong, Guowen Yuan, Rui Shu, Xin Zhou, Xu Han, Yuan Wang, Yun Peng, Zeyang Zhuang, Zihan Wu","submitted_at":"2026-07-07T12:09:46Z","abstract_excerpt":"Coding agents are ranked almost entirely by resolve rate: whether their final patch passes the target tests. Yet two agents can reach the same outcome through very different processes, and a single pass/fail label says nothing about why a run failed or why an accepted run spent extra steps, time, or tokens. This process evidence lives in the trajectory, which records a run's searches, reads, edits, tool calls, validation, and reversions. However, raw traces are heterogeneous and hard to compare across runs. We present TraceProbe, a trajectory-diagnostic framework that recovers what resolve rat"},"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":"2607.06184","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-07-07T12:09:46Z","cross_cats_sorted":[],"title_canon_sha256":"a836a23413e371d277ee70f51c19cf46e88c171919970b070df6404a5c1b15ce","abstract_canon_sha256":"17dff3102cd79968699e9187f7b90b7531897993e0cd9b6adae70be920ffa106"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-08T01:19:16.891141Z","signature_b64":"LOerinAxetmQxsMd958IB3kxMd7BFKfqORXeqU0JT7jUKP3hWmBlnPAk6RdlRXhwon+HakAv1DU3LkKi3nIsAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"56f4301a88a2a5fbe30e7a6da4326aae800c853c1a3ec9f7854c7e4d1fc1a31e","last_reissued_at":"2026-07-08T01:19:16.890518Z","signature_status":"signed_v1","first_computed_at":"2026-07-08T01:19:16.890518Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"What Resolve Rate Hides: Trajectory Structure Diagnostics for Coding Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Chun Yong Chong, Guowen Yuan, Rui Shu, Xin Zhou, Xu Han, Yuan Wang, Yun Peng, Zeyang Zhuang, Zihan Wu","submitted_at":"2026-07-07T12:09:46Z","abstract_excerpt":"Coding agents are ranked almost entirely by resolve rate: whether their final patch passes the target tests. Yet two agents can reach the same outcome through very different processes, and a single pass/fail label says nothing about why a run failed or why an accepted run spent extra steps, time, or tokens. This process evidence lives in the trajectory, which records a run's searches, reads, edits, tool calls, validation, and reversions. However, raw traces are heterogeneous and hard to compare across runs. We present TraceProbe, a trajectory-diagnostic framework that recovers what resolve rat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.06184","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/2607.06184/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":"2607.06184","created_at":"2026-07-08T01:19:16.890585+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.06184v1","created_at":"2026-07-08T01:19:16.890585+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.06184","created_at":"2026-07-08T01:19:16.890585+00:00"},{"alias_kind":"pith_short_12","alias_value":"K32DAGUIUKS7","created_at":"2026-07-08T01:19:16.890585+00:00"},{"alias_kind":"pith_short_16","alias_value":"K32DAGUIUKS7XYYO","created_at":"2026-07-08T01:19:16.890585+00:00"},{"alias_kind":"pith_short_8","alias_value":"K32DAGUI","created_at":"2026-07-08T01:19:16.890585+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/K32DAGUIUKS7XYYOPJW2IMTKV2","json":"https://pith.science/pith/K32DAGUIUKS7XYYOPJW2IMTKV2.json","graph_json":"https://pith.science/api/pith-number/K32DAGUIUKS7XYYOPJW2IMTKV2/graph.json","events_json":"https://pith.science/api/pith-number/K32DAGUIUKS7XYYOPJW2IMTKV2/events.json","paper":"https://pith.science/paper/K32DAGUI"},"agent_actions":{"view_html":"https://pith.science/pith/K32DAGUIUKS7XYYOPJW2IMTKV2","download_json":"https://pith.science/pith/K32DAGUIUKS7XYYOPJW2IMTKV2.json","view_paper":"https://pith.science/paper/K32DAGUI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.06184&json=true","fetch_graph":"https://pith.science/api/pith-number/K32DAGUIUKS7XYYOPJW2IMTKV2/graph.json","fetch_events":"https://pith.science/api/pith-number/K32DAGUIUKS7XYYOPJW2IMTKV2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/K32DAGUIUKS7XYYOPJW2IMTKV2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/K32DAGUIUKS7XYYOPJW2IMTKV2/action/storage_attestation","attest_author":"https://pith.science/pith/K32DAGUIUKS7XYYOPJW2IMTKV2/action/author_attestation","sign_citation":"https://pith.science/pith/K32DAGUIUKS7XYYOPJW2IMTKV2/action/citation_signature","submit_replication":"https://pith.science/pith/K32DAGUIUKS7XYYOPJW2IMTKV2/action/replication_record"}},"created_at":"2026-07-08T01:19:16.890585+00:00","updated_at":"2026-07-08T01:19:16.890585+00:00"}