{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:CEWM4ZL2FC3F6575MTAEME2LS7","short_pith_number":"pith:CEWM4ZL2","schema_version":"1.0","canonical_sha256":"112cce657a28b65f77fd64c046134b97fd3675a321ce1ba31527c9161ada8d9f","source":{"kind":"arxiv","id":"2606.01770","version":1},"attestation_state":"computed","paper":{"title":"Adaptive Auto-Harness: Sustained Self-Improvement for Agentic System Deployment on Open-Ended Task Streams","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Benoit Dumoulin, Bing He, Dakuo Wang, Hanqing Lu, Minhua Lin, Tianxin Wei, Wei Jin, Yisi Sang, Zewen Liu, Zhan Shi","submitted_at":"2026-06-01T06:51:14Z","abstract_excerpt":"Auto-harness systems such as A-Evolve, GEPA, and Meta-Harness improve LLM agents by optimizing prompts, skills, tools, memories, and supporting infrastructure from execution feedback, but they are typically evaluated on fixed offline benchmarks. Real deployments instead present open-ended task streams: histories grow without a fixed endpoint, heterogeneous tasks require different harnesses, and problem distributions shift over time. These challenges make a single repeatedly and densely updated harness brittle, causing performance degradation as accuracy peaks early and then declines. This moti"},"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":"2606.01770","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-01T06:51:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"31e92c8e9fcd5972ad8da823b4d60bc1ed054224b69a33a6fca9b17e3acb069c","abstract_canon_sha256":"d2178dc4d29efdb048efa017be94fd1d0392a7cc5c6b3234dfe23010b193274e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:56.178165Z","signature_b64":"eh2U+CXwhQcnYQcEokKGY2fKoketMQTFuhJm62Q2GShisEGCYFDihs9XV7vfyjRqePouGu/q0+b+UKOurO2vBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"112cce657a28b65f77fd64c046134b97fd3675a321ce1ba31527c9161ada8d9f","last_reissued_at":"2026-06-02T02:04:56.177808Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:56.177808Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Adaptive Auto-Harness: Sustained Self-Improvement for Agentic System Deployment on Open-Ended Task Streams","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Benoit Dumoulin, Bing He, Dakuo Wang, Hanqing Lu, Minhua Lin, Tianxin Wei, Wei Jin, Yisi Sang, Zewen Liu, Zhan Shi","submitted_at":"2026-06-01T06:51:14Z","abstract_excerpt":"Auto-harness systems such as A-Evolve, GEPA, and Meta-Harness improve LLM agents by optimizing prompts, skills, tools, memories, and supporting infrastructure from execution feedback, but they are typically evaluated on fixed offline benchmarks. Real deployments instead present open-ended task streams: histories grow without a fixed endpoint, heterogeneous tasks require different harnesses, and problem distributions shift over time. These challenges make a single repeatedly and densely updated harness brittle, causing performance degradation as accuracy peaks early and then declines. This moti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.01770","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/2606.01770/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":"2606.01770","created_at":"2026-06-02T02:04:56.177863+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.01770v1","created_at":"2026-06-02T02:04:56.177863+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.01770","created_at":"2026-06-02T02:04:56.177863+00:00"},{"alias_kind":"pith_short_12","alias_value":"CEWM4ZL2FC3F","created_at":"2026-06-02T02:04:56.177863+00:00"},{"alias_kind":"pith_short_16","alias_value":"CEWM4ZL2FC3F6575","created_at":"2026-06-02T02:04:56.177863+00:00"},{"alias_kind":"pith_short_8","alias_value":"CEWM4ZL2","created_at":"2026-06-02T02:04:56.177863+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/CEWM4ZL2FC3F6575MTAEME2LS7","json":"https://pith.science/pith/CEWM4ZL2FC3F6575MTAEME2LS7.json","graph_json":"https://pith.science/api/pith-number/CEWM4ZL2FC3F6575MTAEME2LS7/graph.json","events_json":"https://pith.science/api/pith-number/CEWM4ZL2FC3F6575MTAEME2LS7/events.json","paper":"https://pith.science/paper/CEWM4ZL2"},"agent_actions":{"view_html":"https://pith.science/pith/CEWM4ZL2FC3F6575MTAEME2LS7","download_json":"https://pith.science/pith/CEWM4ZL2FC3F6575MTAEME2LS7.json","view_paper":"https://pith.science/paper/CEWM4ZL2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.01770&json=true","fetch_graph":"https://pith.science/api/pith-number/CEWM4ZL2FC3F6575MTAEME2LS7/graph.json","fetch_events":"https://pith.science/api/pith-number/CEWM4ZL2FC3F6575MTAEME2LS7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CEWM4ZL2FC3F6575MTAEME2LS7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CEWM4ZL2FC3F6575MTAEME2LS7/action/storage_attestation","attest_author":"https://pith.science/pith/CEWM4ZL2FC3F6575MTAEME2LS7/action/author_attestation","sign_citation":"https://pith.science/pith/CEWM4ZL2FC3F6575MTAEME2LS7/action/citation_signature","submit_replication":"https://pith.science/pith/CEWM4ZL2FC3F6575MTAEME2LS7/action/replication_record"}},"created_at":"2026-06-02T02:04:56.177863+00:00","updated_at":"2026-06-02T02:04:56.177863+00:00"}