{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:DD5GZWMN45FUVV23DVPRS2HGCY","short_pith_number":"pith:DD5GZWMN","canonical_record":{"source":{"id":"2607.16345","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-07-16T21:33:05Z","cross_cats_sorted":["cs.AI","cs.LG","cs.PF"],"title_canon_sha256":"e609518a7dd3910b72c8a29e503c03d432b88211ef58519758dce57b25c2ebff","abstract_canon_sha256":"dea8cc53910dab386251519b32f1cbd0cf8fae60e66cd540edeff7a8d0e79643"},"schema_version":"1.0"},"canonical_sha256":"18fa6cd98de74b4ad75b1d5f1968e6160b1eefdd90aaa4110a676d62b768fa31","source":{"kind":"arxiv","id":"2607.16345","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.16345","created_at":"2026-07-21T00:20:13Z"},{"alias_kind":"arxiv_version","alias_value":"2607.16345v1","created_at":"2026-07-21T00:20:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.16345","created_at":"2026-07-21T00:20:13Z"},{"alias_kind":"pith_short_12","alias_value":"DD5GZWMN45FU","created_at":"2026-07-21T00:20:13Z"},{"alias_kind":"pith_short_16","alias_value":"DD5GZWMN45FUVV23","created_at":"2026-07-21T00:20:13Z"},{"alias_kind":"pith_short_8","alias_value":"DD5GZWMN","created_at":"2026-07-21T00:20:13Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:DD5GZWMN45FUVV23DVPRS2HGCY","target":"record","payload":{"canonical_record":{"source":{"id":"2607.16345","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-07-16T21:33:05Z","cross_cats_sorted":["cs.AI","cs.LG","cs.PF"],"title_canon_sha256":"e609518a7dd3910b72c8a29e503c03d432b88211ef58519758dce57b25c2ebff","abstract_canon_sha256":"dea8cc53910dab386251519b32f1cbd0cf8fae60e66cd540edeff7a8d0e79643"},"schema_version":"1.0"},"canonical_sha256":"18fa6cd98de74b4ad75b1d5f1968e6160b1eefdd90aaa4110a676d62b768fa31","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-21T00:20:13.434689Z","signature_b64":"PpBJmm5PIMtZMoUaDr55pxXj0wJnQuYYP+jZziRD9nTwvn0vrZFcKk/l2QDKK8bGcFdhoHmPnYAXUkNH8zW0Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"18fa6cd98de74b4ad75b1d5f1968e6160b1eefdd90aaa4110a676d62b768fa31","last_reissued_at":"2026-07-21T00:20:13.433845Z","signature_status":"signed_v1","first_computed_at":"2026-07-21T00:20:13.433845Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.16345","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":"2026-07-21T00:20:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uzvaT4VOLIXae5zQIJWIf5chHssL89yfqmoVb+TfKLjXw/AEvsfcry7nIZe9cqo45O/7PjTR6wE5HXjJSOIbCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T10:45:26.387149Z"},"content_sha256":"e8a90df2d0f074cd4f02a91357786fc8a095f0c070b125f89fe4d69e9635a853","schema_version":"1.0","event_id":"sha256:e8a90df2d0f074cd4f02a91357786fc8a095f0c070b125f89fe4d69e9635a853"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:DD5GZWMN45FUVV23DVPRS2HGCY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"AEVAL: From Anecdotal to Deterministic Testing for Agentic Skill Workflows","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.PF"],"primary_cat":"cs.SE","authors_text":"Bingjie Zhou, Steve Masson, Tejas Singh Anand, Tian Zheng, Wanting Jiang, Yuet Ying Christina Wang","submitted_at":"2026-07-16T21:33:05Z","abstract_excerpt":"Modern agentic systems increasingly rely on skills: installable packages of natural language and code that teach an LLM agent to perform a domain task. As skill repositories grow, developers need automated quality signals on every change, yet evaluation today is largely anecdotal: a developer asks an agent to \"try the skill,\" watches a demo, and forms a subjective impression. This yields neither reproducibility across runs nor comparability across versions, and scales poorly to marketplaces where one regression can silently break dozens of downstream workflows.\n  We present AEVAL (Agentic Eval"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.16345","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/2607.16345/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":"2026-07-21T00:20:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bJ8L+fQIHrSGAQ63X8vjmuGAvWE9J/ShUbw/1SSNiACZ5eJhQXo3K5wTaG8pfl9pNu0A3s7vOOJy9fKV27S1BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T10:45:26.387719Z"},"content_sha256":"b8c859cce147115ada4691ded7dbb66069e1c61b173da6e08c7835f57587e489","schema_version":"1.0","event_id":"sha256:b8c859cce147115ada4691ded7dbb66069e1c61b173da6e08c7835f57587e489"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DD5GZWMN45FUVV23DVPRS2HGCY/bundle.json","state_url":"https://pith.science/pith/DD5GZWMN45FUVV23DVPRS2HGCY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DD5GZWMN45FUVV23DVPRS2HGCY/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":"2026-08-15T10:45:26Z","links":{"resolver":"https://pith.science/pith/DD5GZWMN45FUVV23DVPRS2HGCY","bundle":"https://pith.science/pith/DD5GZWMN45FUVV23DVPRS2HGCY/bundle.json","state":"https://pith.science/pith/DD5GZWMN45FUVV23DVPRS2HGCY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DD5GZWMN45FUVV23DVPRS2HGCY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:DD5GZWMN45FUVV23DVPRS2HGCY","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":"dea8cc53910dab386251519b32f1cbd0cf8fae60e66cd540edeff7a8d0e79643","cross_cats_sorted":["cs.AI","cs.LG","cs.PF"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-07-16T21:33:05Z","title_canon_sha256":"e609518a7dd3910b72c8a29e503c03d432b88211ef58519758dce57b25c2ebff"},"schema_version":"1.0","source":{"id":"2607.16345","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.16345","created_at":"2026-07-21T00:20:13Z"},{"alias_kind":"arxiv_version","alias_value":"2607.16345v1","created_at":"2026-07-21T00:20:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.16345","created_at":"2026-07-21T00:20:13Z"},{"alias_kind":"pith_short_12","alias_value":"DD5GZWMN45FU","created_at":"2026-07-21T00:20:13Z"},{"alias_kind":"pith_short_16","alias_value":"DD5GZWMN45FUVV23","created_at":"2026-07-21T00:20:13Z"},{"alias_kind":"pith_short_8","alias_value":"DD5GZWMN","created_at":"2026-07-21T00:20:13Z"}],"graph_snapshots":[{"event_id":"sha256:b8c859cce147115ada4691ded7dbb66069e1c61b173da6e08c7835f57587e489","target":"graph","created_at":"2026-07-21T00:20:13Z","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/2607.16345/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Modern agentic systems increasingly rely on skills: installable packages of natural language and code that teach an LLM agent to perform a domain task. As skill repositories grow, developers need automated quality signals on every change, yet evaluation today is largely anecdotal: a developer asks an agent to \"try the skill,\" watches a demo, and forms a subjective impression. This yields neither reproducibility across runs nor comparability across versions, and scales poorly to marketplaces where one regression can silently break dozens of downstream workflows.\n  We present AEVAL (Agentic Eval","authors_text":"Bingjie Zhou, Steve Masson, Tejas Singh Anand, Tian Zheng, Wanting Jiang, Yuet Ying Christina Wang","cross_cats":["cs.AI","cs.LG","cs.PF"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-07-16T21:33:05Z","title":"AEVAL: From Anecdotal to Deterministic Testing for Agentic Skill Workflows"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.16345","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:e8a90df2d0f074cd4f02a91357786fc8a095f0c070b125f89fe4d69e9635a853","target":"record","created_at":"2026-07-21T00:20:13Z","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":"dea8cc53910dab386251519b32f1cbd0cf8fae60e66cd540edeff7a8d0e79643","cross_cats_sorted":["cs.AI","cs.LG","cs.PF"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-07-16T21:33:05Z","title_canon_sha256":"e609518a7dd3910b72c8a29e503c03d432b88211ef58519758dce57b25c2ebff"},"schema_version":"1.0","source":{"id":"2607.16345","kind":"arxiv","version":1}},"canonical_sha256":"18fa6cd98de74b4ad75b1d5f1968e6160b1eefdd90aaa4110a676d62b768fa31","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"18fa6cd98de74b4ad75b1d5f1968e6160b1eefdd90aaa4110a676d62b768fa31","first_computed_at":"2026-07-21T00:20:13.433845Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-21T00:20:13.433845Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"PpBJmm5PIMtZMoUaDr55pxXj0wJnQuYYP+jZziRD9nTwvn0vrZFcKk/l2QDKK8bGcFdhoHmPnYAXUkNH8zW0Cg==","signature_status":"signed_v1","signed_at":"2026-07-21T00:20:13.434689Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.16345","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e8a90df2d0f074cd4f02a91357786fc8a095f0c070b125f89fe4d69e9635a853","sha256:b8c859cce147115ada4691ded7dbb66069e1c61b173da6e08c7835f57587e489"],"state_sha256":"4471edcbd9cbfb0e2911c8159ba5abc5a89de3eb627a683b212a1688d838c563"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"vm2cbYxL3AyaOxNnl2slHzN4YqgPVO0i+ItliMnsP3Fy598jMrx4VNCior8xoZ2RmCJ2LeXPqC8uQ6OlilqHAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-15T10:45:26.392949Z","bundle_sha256":"4bcb646662d9c8be4acb1e14b54dde7b3fc47432fb217d7f26b81c98b69a27cd"}}