{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:ART6LZYZDKVCL23Z2EXLHROG4J","short_pith_number":"pith:ART6LZYZ","canonical_record":{"source":{"id":"2603.26337","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-03-27T11:58:47Z","cross_cats_sorted":[],"title_canon_sha256":"ccb8df41a866f7b8224f94ca0ad7afa46ae9357c2ffcc91e6329475a0028fab1","abstract_canon_sha256":"a31e1eb723e0a69752e3341d19349948826b2265ecd8dd2483424b01d5f1d792"},"schema_version":"1.0"},"canonical_sha256":"0467e5e7191aaa25eb79d12eb3c5c6e26f95a94e532d92cc30085d89984c1aef","source":{"kind":"arxiv","id":"2603.26337","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.26337","created_at":"2026-08-07T00:51:32Z"},{"alias_kind":"arxiv_version","alias_value":"2603.26337v2","created_at":"2026-08-07T00:51:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.26337","created_at":"2026-08-07T00:51:32Z"},{"alias_kind":"pith_short_12","alias_value":"ART6LZYZDKVC","created_at":"2026-08-07T00:51:32Z"},{"alias_kind":"pith_short_16","alias_value":"ART6LZYZDKVCL23Z","created_at":"2026-08-07T00:51:32Z"},{"alias_kind":"pith_short_8","alias_value":"ART6LZYZ","created_at":"2026-08-07T00:51:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:ART6LZYZDKVCL23Z2EXLHROG4J","target":"record","payload":{"canonical_record":{"source":{"id":"2603.26337","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-03-27T11:58:47Z","cross_cats_sorted":[],"title_canon_sha256":"ccb8df41a866f7b8224f94ca0ad7afa46ae9357c2ffcc91e6329475a0028fab1","abstract_canon_sha256":"a31e1eb723e0a69752e3341d19349948826b2265ecd8dd2483424b01d5f1d792"},"schema_version":"1.0"},"canonical_sha256":"0467e5e7191aaa25eb79d12eb3c5c6e26f95a94e532d92cc30085d89984c1aef","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-07T00:51:32.081237Z","signature_b64":"35XjHfdX2LaenrSi3PU4gXgRNLwL0fIeJZoiL6wzj0WS9xUKx93C/zblCZNR6jbqduij6uI9EIfQIjhFH2kEAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0467e5e7191aaa25eb79d12eb3c5c6e26f95a94e532d92cc30085d89984c1aef","last_reissued_at":"2026-08-07T00:51:32.079583Z","signature_status":"signed_v1","first_computed_at":"2026-08-07T00:51:32.079583Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2603.26337","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-08-07T00:51:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"slVF3Nb4kBVXqG6DZrkb+8s/b+Dggto/iQ++6GYxqYcwz5178e8QKRn8CIzYNCuQgASdsOVbNCynzUkx16DnAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T19:16:08.780478Z"},"content_sha256":"f9957db5af85020e3bba32b1d9b9d4739c7c604731fb3c97d73066bad9b81eb0","schema_version":"1.0","event_id":"sha256:f9957db5af85020e3bba32b1d9b9d4739c7c604731fb3c97d73066bad9b81eb0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:ART6LZYZDKVCL23Z2EXLHROG4J","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"A Benchmark for Evaluating Repository-Level Code Agents with Intermediate Reasoning on Feature Addition Task","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Kui Liu, Shuhan Liu, Xiaohu Yang, Xing Hu, Xin Xia, Zhiyi Zhao","submitted_at":"2026-03-27T11:58:47Z","abstract_excerpt":"Repository-level code agents have shown strong promise in real-world feature addition tasks, making reliable evaluation of their capabilities increasingly important. However, existing benchmarks primarily evaluate these agents as black boxes based on final test correctness, providing limited insight into how they reason and where failures arise. To address this limitation, we introduce RACE-bench, a reasoning-augmented benchmark for evaluating code agents on repository-level feature addition tasks. RACE-bench contains 528 real-world feature addition instances from 12 open-source repositories. "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.26337","kind":"arxiv","version":2},"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/2603.26337/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":"2026-08-07T00:51:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"s5wGgHfbpRk8ekW3lxwkix+vf6t58r4OTwUXhd3CR1plCZrsrcAWo3IBdubS3HmypVQB7j2sSlHs60ionWyuAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T19:16:08.781282Z"},"content_sha256":"dfae1275b0bf8ebf7a9ea2f657b3eef2cd97853712c42290cdf99e736078aec0","schema_version":"1.0","event_id":"sha256:dfae1275b0bf8ebf7a9ea2f657b3eef2cd97853712c42290cdf99e736078aec0"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ART6LZYZDKVCL23Z2EXLHROG4J/bundle.json","state_url":"https://pith.science/pith/ART6LZYZDKVCL23Z2EXLHROG4J/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ART6LZYZDKVCL23Z2EXLHROG4J/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":"2026-08-19T19:16:08Z","links":{"resolver":"https://pith.science/pith/ART6LZYZDKVCL23Z2EXLHROG4J","bundle":"https://pith.science/pith/ART6LZYZDKVCL23Z2EXLHROG4J/bundle.json","state":"https://pith.science/pith/ART6LZYZDKVCL23Z2EXLHROG4J/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ART6LZYZDKVCL23Z2EXLHROG4J/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:ART6LZYZDKVCL23Z2EXLHROG4J","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":"a31e1eb723e0a69752e3341d19349948826b2265ecd8dd2483424b01d5f1d792","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-03-27T11:58:47Z","title_canon_sha256":"ccb8df41a866f7b8224f94ca0ad7afa46ae9357c2ffcc91e6329475a0028fab1"},"schema_version":"1.0","source":{"id":"2603.26337","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.26337","created_at":"2026-08-07T00:51:32Z"},{"alias_kind":"arxiv_version","alias_value":"2603.26337v2","created_at":"2026-08-07T00:51:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.26337","created_at":"2026-08-07T00:51:32Z"},{"alias_kind":"pith_short_12","alias_value":"ART6LZYZDKVC","created_at":"2026-08-07T00:51:32Z"},{"alias_kind":"pith_short_16","alias_value":"ART6LZYZDKVCL23Z","created_at":"2026-08-07T00:51:32Z"},{"alias_kind":"pith_short_8","alias_value":"ART6LZYZ","created_at":"2026-08-07T00:51:32Z"}],"graph_snapshots":[{"event_id":"sha256:dfae1275b0bf8ebf7a9ea2f657b3eef2cd97853712c42290cdf99e736078aec0","target":"graph","created_at":"2026-08-07T00:51:32Z","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/2603.26337/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Repository-level code agents have shown strong promise in real-world feature addition tasks, making reliable evaluation of their capabilities increasingly important. However, existing benchmarks primarily evaluate these agents as black boxes based on final test correctness, providing limited insight into how they reason and where failures arise. To address this limitation, we introduce RACE-bench, a reasoning-augmented benchmark for evaluating code agents on repository-level feature addition tasks. RACE-bench contains 528 real-world feature addition instances from 12 open-source repositories. ","authors_text":"Kui Liu, Shuhan Liu, Xiaohu Yang, Xing Hu, Xin Xia, Zhiyi Zhao","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-03-27T11:58:47Z","title":"A Benchmark for Evaluating Repository-Level Code Agents with Intermediate Reasoning on Feature Addition Task"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.26337","kind":"arxiv","version":2},"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:f9957db5af85020e3bba32b1d9b9d4739c7c604731fb3c97d73066bad9b81eb0","target":"record","created_at":"2026-08-07T00:51:32Z","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":"a31e1eb723e0a69752e3341d19349948826b2265ecd8dd2483424b01d5f1d792","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-03-27T11:58:47Z","title_canon_sha256":"ccb8df41a866f7b8224f94ca0ad7afa46ae9357c2ffcc91e6329475a0028fab1"},"schema_version":"1.0","source":{"id":"2603.26337","kind":"arxiv","version":2}},"canonical_sha256":"0467e5e7191aaa25eb79d12eb3c5c6e26f95a94e532d92cc30085d89984c1aef","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0467e5e7191aaa25eb79d12eb3c5c6e26f95a94e532d92cc30085d89984c1aef","first_computed_at":"2026-08-07T00:51:32.079583Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-07T00:51:32.079583Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"35XjHfdX2LaenrSi3PU4gXgRNLwL0fIeJZoiL6wzj0WS9xUKx93C/zblCZNR6jbqduij6uI9EIfQIjhFH2kEAw==","signature_status":"signed_v1","signed_at":"2026-08-07T00:51:32.081237Z","signed_message":"canonical_sha256_bytes"},"source_id":"2603.26337","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f9957db5af85020e3bba32b1d9b9d4739c7c604731fb3c97d73066bad9b81eb0","sha256:dfae1275b0bf8ebf7a9ea2f657b3eef2cd97853712c42290cdf99e736078aec0"],"state_sha256":"d88706c7981e40c929eb464ab4bb6c02e090a3badd0a77ec3e328cf0eee7fbd9"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RAP5rkQYzTbCmw37yAJ+lbrgRIFyzpajgbhPH+5s6IUNFBGPR4s7bw29UupA0oqowGxyOS/r0MqHG8MajqZoDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-19T19:16:08.788310Z","bundle_sha256":"3cef1c3768b35a5fa6d036535c173af2db8ad1a4dcbb6955af56a707f13742b3"}}