{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:L4MWM4EANSVTOFJZOKGPJ64SFF","short_pith_number":"pith:L4MWM4EA","canonical_record":{"source":{"id":"2607.08124","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-07-09T05:53:39Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"5c07bfc9375a4189f858c406e0bdd391fb53310cba3095d5b9abff9962cb7a70","abstract_canon_sha256":"4f851ade397cf828e16256bfc4e66fdd6e0df9e41db7c6406ce484f8820a19ee"},"schema_version":"1.0"},"canonical_sha256":"5f196670806cab371539728cf4fb92294d55bedfbcd975112ec5a4e0bce3ec1f","source":{"kind":"arxiv","id":"2607.08124","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.08124","created_at":"2026-07-10T01:19:27Z"},{"alias_kind":"arxiv_version","alias_value":"2607.08124v1","created_at":"2026-07-10T01:19:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.08124","created_at":"2026-07-10T01:19:27Z"},{"alias_kind":"pith_short_12","alias_value":"L4MWM4EANSVT","created_at":"2026-07-10T01:19:27Z"},{"alias_kind":"pith_short_16","alias_value":"L4MWM4EANSVTOFJZ","created_at":"2026-07-10T01:19:27Z"},{"alias_kind":"pith_short_8","alias_value":"L4MWM4EA","created_at":"2026-07-10T01:19:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:L4MWM4EANSVTOFJZOKGPJ64SFF","target":"record","payload":{"canonical_record":{"source":{"id":"2607.08124","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-07-09T05:53:39Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"5c07bfc9375a4189f858c406e0bdd391fb53310cba3095d5b9abff9962cb7a70","abstract_canon_sha256":"4f851ade397cf828e16256bfc4e66fdd6e0df9e41db7c6406ce484f8820a19ee"},"schema_version":"1.0"},"canonical_sha256":"5f196670806cab371539728cf4fb92294d55bedfbcd975112ec5a4e0bce3ec1f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-10T01:19:27.328682Z","signature_b64":"TYwDGBE9LgNBT/j6KdVlswaJQI+1cfcdReUbxhwEYFFWCz+0uWnV0NBFNaof8w4k04xkrmZpyzcCbmlaY9O2Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5f196670806cab371539728cf4fb92294d55bedfbcd975112ec5a4e0bce3ec1f","last_reissued_at":"2026-07-10T01:19:27.328186Z","signature_status":"signed_v1","first_computed_at":"2026-07-10T01:19:27.328186Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.08124","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":"2026-07-10T01:19:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"88h4D+w6I5BW6MeFcK9IpsTvHxfYVLGGe8DT981KDQXGkkYnNSbhoYvhb2f5N93cNtXOKDCHZkYA/XMcpqsQAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-22T14:11:25.822073Z"},"content_sha256":"6a486c7790acb743ffbcc8382a09eb83cbd62b66ef057355e9f1b9df71039c85","schema_version":"1.0","event_id":"sha256:6a486c7790acb743ffbcc8382a09eb83cbd62b66ef057355e9f1b9df71039c85"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:L4MWM4EANSVTOFJZOKGPJ64SFF","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"TTHE: Test-Time Harness Evolution","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.SE","authors_text":"Bo Han, Dahai Yu, Jun Nie, Jun Song, Qianshu Cai, Xinmei Tian, Yike Guo, Yonggang Zhang","submitted_at":"2026-07-09T05:53:39Z","abstract_excerpt":"The behavior of an LLM agent is determined not only by the underlying model, but also by its harness: the executable program that constructs context, invokes tools, verifies intermediate results, and recovers from failures. Existing approaches optimize such harnesses before deployment, searching training or development data for a fixed agent workflow that is then frozen at test time. This limits adaptation when the test distribution, failure modes, or tool interactions differ from those seen during development. We ask whether the harness can instead be optimized during evaluation itself, using"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.08124","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/2607.08124/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":"2026-07-10T01:19:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EyKdRpkYXOw/W7rrX7kDpysbsb38vam8cI0w+QR8soNGYDfd6w+WqlsUOgEzcjQpV0iXCS4JyKtJQeJqDKbRDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-22T14:11:25.822461Z"},"content_sha256":"4e823af2e57cda21eecd7ef2753a854f8b88f84803d3e05787dfc88dc354ea61","schema_version":"1.0","event_id":"sha256:4e823af2e57cda21eecd7ef2753a854f8b88f84803d3e05787dfc88dc354ea61"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/L4MWM4EANSVTOFJZOKGPJ64SFF/bundle.json","state_url":"https://pith.science/pith/L4MWM4EANSVTOFJZOKGPJ64SFF/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/L4MWM4EANSVTOFJZOKGPJ64SFF/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":"2026-07-22T14:11:25Z","links":{"resolver":"https://pith.science/pith/L4MWM4EANSVTOFJZOKGPJ64SFF","bundle":"https://pith.science/pith/L4MWM4EANSVTOFJZOKGPJ64SFF/bundle.json","state":"https://pith.science/pith/L4MWM4EANSVTOFJZOKGPJ64SFF/state.json","well_known_bundle":"https://pith.science/.well-known/pith/L4MWM4EANSVTOFJZOKGPJ64SFF/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:L4MWM4EANSVTOFJZOKGPJ64SFF","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":"4f851ade397cf828e16256bfc4e66fdd6e0df9e41db7c6406ce484f8820a19ee","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-07-09T05:53:39Z","title_canon_sha256":"5c07bfc9375a4189f858c406e0bdd391fb53310cba3095d5b9abff9962cb7a70"},"schema_version":"1.0","source":{"id":"2607.08124","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.08124","created_at":"2026-07-10T01:19:27Z"},{"alias_kind":"arxiv_version","alias_value":"2607.08124v1","created_at":"2026-07-10T01:19:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.08124","created_at":"2026-07-10T01:19:27Z"},{"alias_kind":"pith_short_12","alias_value":"L4MWM4EANSVT","created_at":"2026-07-10T01:19:27Z"},{"alias_kind":"pith_short_16","alias_value":"L4MWM4EANSVTOFJZ","created_at":"2026-07-10T01:19:27Z"},{"alias_kind":"pith_short_8","alias_value":"L4MWM4EA","created_at":"2026-07-10T01:19:27Z"}],"graph_snapshots":[{"event_id":"sha256:4e823af2e57cda21eecd7ef2753a854f8b88f84803d3e05787dfc88dc354ea61","target":"graph","created_at":"2026-07-10T01:19:27Z","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/2607.08124/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The behavior of an LLM agent is determined not only by the underlying model, but also by its harness: the executable program that constructs context, invokes tools, verifies intermediate results, and recovers from failures. Existing approaches optimize such harnesses before deployment, searching training or development data for a fixed agent workflow that is then frozen at test time. This limits adaptation when the test distribution, failure modes, or tool interactions differ from those seen during development. We ask whether the harness can instead be optimized during evaluation itself, using","authors_text":"Bo Han, Dahai Yu, Jun Nie, Jun Song, Qianshu Cai, Xinmei Tian, Yike Guo, Yonggang Zhang","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-07-09T05:53:39Z","title":"TTHE: Test-Time Harness Evolution"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.08124","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":"sha256:6a486c7790acb743ffbcc8382a09eb83cbd62b66ef057355e9f1b9df71039c85","target":"record","created_at":"2026-07-10T01:19:27Z","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":"4f851ade397cf828e16256bfc4e66fdd6e0df9e41db7c6406ce484f8820a19ee","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-07-09T05:53:39Z","title_canon_sha256":"5c07bfc9375a4189f858c406e0bdd391fb53310cba3095d5b9abff9962cb7a70"},"schema_version":"1.0","source":{"id":"2607.08124","kind":"arxiv","version":1}},"canonical_sha256":"5f196670806cab371539728cf4fb92294d55bedfbcd975112ec5a4e0bce3ec1f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5f196670806cab371539728cf4fb92294d55bedfbcd975112ec5a4e0bce3ec1f","first_computed_at":"2026-07-10T01:19:27.328186Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-10T01:19:27.328186Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"TYwDGBE9LgNBT/j6KdVlswaJQI+1cfcdReUbxhwEYFFWCz+0uWnV0NBFNaof8w4k04xkrmZpyzcCbmlaY9O2Bg==","signature_status":"signed_v1","signed_at":"2026-07-10T01:19:27.328682Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.08124","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6a486c7790acb743ffbcc8382a09eb83cbd62b66ef057355e9f1b9df71039c85","sha256:4e823af2e57cda21eecd7ef2753a854f8b88f84803d3e05787dfc88dc354ea61"],"state_sha256":"2ea62e2a2f1c564ce35d8c7328773ff8a98201846e1c5bc403a7caf2af10500b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SHrUiW0XQh8H3P5CdZ1ayHkvxD/6CDI8yaFA+X1+EXSRvqhyvUwhWrcvQr5rMnP8z6wYnOPVJfTRNEVt6W7sBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-22T14:11:25.824987Z","bundle_sha256":"59d6e1c10449c25c4ba346e915dab0be8f40e802062a78b5bb41aa1e8926bc9d"}}