{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:BNUTSSXBFV36YQYX4MO53WGSXF","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":"1556b25725f62e7978320a558c33b0966f1c651c7bf2b8b244ee41259ea88f0f","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-03T08:51:50Z","title_canon_sha256":"183e2977f28a8c9526d04e2fcab9c1119083b962f7b43713ff9fc123cf8057c2"},"schema_version":"1.0","source":{"id":"2608.01918","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.01918","created_at":"2026-08-04T02:07:42Z"},{"alias_kind":"arxiv_version","alias_value":"2608.01918v1","created_at":"2026-08-04T02:07:42Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.01918","created_at":"2026-08-04T02:07:42Z"},{"alias_kind":"pith_short_12","alias_value":"BNUTSSXBFV36","created_at":"2026-08-04T02:07:42Z"},{"alias_kind":"pith_short_16","alias_value":"BNUTSSXBFV36YQYX","created_at":"2026-08-04T02:07:42Z"},{"alias_kind":"pith_short_8","alias_value":"BNUTSSXB","created_at":"2026-08-04T02:07:42Z"}],"graph_snapshots":[{"event_id":"sha256:a43b1a7bad1118e1a40ea29c3ba472e04ad8c3133668733b3c753c13072f9fce","target":"graph","created_at":"2026-08-04T02:07:42Z","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/2608.01918/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Harness design plays a critical role in agent performance by shaping how large language models (LLMs) perceive, reason over, and act within executable environments. Recent work has proposed automatic harness evolution, which iteratively improves the harness from agent--environment interactions. However, existing methods often overfit to the evolution tasks, rely exclusively on trajectory-derived signals, and optimize harness components jointly, causing interference across components. We propose HarnessCompass, a novel automatic harness evolution framework built around constrained evolution, pr","authors_text":"Chenhao Li, Dandan Song, Guangyuan Feng, Huipeng Ma, Jun Yang, Luan Zhang, Ruochen Zhou, Xudong Li, Yan Xu, Yizhou Jin, Yuhang Tian, Zhengyu Chen","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-03T08:51:50Z","title":"HarnessCompass: Guiding Automatic Harness Evolution toward Generalizable and Effective Agent Harnesses"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.01918","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":"sha256:1e23e5d9ac4bfa8d75064be59f861ef89a51d98da1248bda1e6b82c04f2743e9","target":"record","created_at":"2026-08-04T02:07:42Z","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":"1556b25725f62e7978320a558c33b0966f1c651c7bf2b8b244ee41259ea88f0f","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-03T08:51:50Z","title_canon_sha256":"183e2977f28a8c9526d04e2fcab9c1119083b962f7b43713ff9fc123cf8057c2"},"schema_version":"1.0","source":{"id":"2608.01918","kind":"arxiv","version":1}},"canonical_sha256":"0b69394ae12d77ec4317e31dddd8d2b96bf5b9bc5c47a0336dd32c557c3eddc1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0b69394ae12d77ec4317e31dddd8d2b96bf5b9bc5c47a0336dd32c557c3eddc1","first_computed_at":"2026-08-04T02:07:42.017644Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-04T02:07:42.017644Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZvCY3+qJxvBOGWFvJqKITpNvXkzFA7MihT5lP5fS/BYurRN9mcNmO/gG2lR7R6PYBDHmFWKpF7Pz3Lb+mu4XBw==","signature_status":"signed_v1","signed_at":"2026-08-04T02:07:42.019169Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.01918","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1e23e5d9ac4bfa8d75064be59f861ef89a51d98da1248bda1e6b82c04f2743e9","sha256:a43b1a7bad1118e1a40ea29c3ba472e04ad8c3133668733b3c753c13072f9fce"],"state_sha256":"7e37ca070fb35cf7bccae233eb8e58e57a86799be4e096dcc27ce42cf9477fdc"}