{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:5K53UDX6QNRFDJ4TL5KM7Z2Y7I","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":"325a977d04776de69d549e0771ca519b851b00e2a9da34503d0511e87dd3534f","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-03-06T04:56:18Z","title_canon_sha256":"47ca4f3bf33cd76be5962b42017385a4893e6a8fb7c84cec9bae736adce9fac2"},"schema_version":"1.0","source":{"id":"2603.05910","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.05910","created_at":"2026-05-20T02:05:41Z"},{"alias_kind":"arxiv_version","alias_value":"2603.05910v2","created_at":"2026-05-20T02:05:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.05910","created_at":"2026-05-20T02:05:41Z"},{"alias_kind":"pith_short_12","alias_value":"5K53UDX6QNRF","created_at":"2026-05-20T02:05:41Z"},{"alias_kind":"pith_short_16","alias_value":"5K53UDX6QNRFDJ4T","created_at":"2026-05-20T02:05:41Z"},{"alias_kind":"pith_short_8","alias_value":"5K53UDX6","created_at":"2026-05-20T02:05:41Z"}],"graph_snapshots":[{"event_id":"sha256:35e839682a012419622d745dcbe3d3bd7e5319be4a8fc9a254075d2271a76fdd","target":"graph","created_at":"2026-05-20T02:05:41Z","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/2603.05910/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"LLM-powered tool-calling agents fulfill user requests by interacting with environments, querying data, and invoking tools in a multi-turn process. Yet, most existing benchmarks evaluate these systems under static environment interfaces, with fixed schemas and toolsets, making it difficult to assess how agents behave as environments evolves -- when capabilities are added, reorganized, or deprecated across successive environment versions. In this paper, we study structured environment evolution as a benchmark-construction problem for tool-calling agents. We propose ProEvolve, a graph-based frame","authors_text":"Bingrou Zhou, Binit Jha, Chen Luo, Dawn Song, Guangrui Li, Jason Choi, Michael J. Morais, Monica Xiao Cheng, Shaunak Mishra, Tianqi Zheng, Xingyuan Pan, Yaochen Xie, Yi Liu, Ziwei Dong","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-03-06T04:56:18Z","title":"The World Won't Stay Still: Programmable Evolution for Agent Benchmarks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.05910","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":"sha256:11805b2d5bbcf3c9f90404f2f80da0a1671510e8d6d3d2e86a538431a35051a2","target":"record","created_at":"2026-05-20T02:05:41Z","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":"325a977d04776de69d549e0771ca519b851b00e2a9da34503d0511e87dd3534f","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-03-06T04:56:18Z","title_canon_sha256":"47ca4f3bf33cd76be5962b42017385a4893e6a8fb7c84cec9bae736adce9fac2"},"schema_version":"1.0","source":{"id":"2603.05910","kind":"arxiv","version":2}},"canonical_sha256":"eabbba0efe836251a7935f54cfe758fa1f5a55765a3886d4bcb68b179aeda599","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"eabbba0efe836251a7935f54cfe758fa1f5a55765a3886d4bcb68b179aeda599","first_computed_at":"2026-05-20T02:05:41.639197Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T02:05:41.639197Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"OWsnNWBRL1rfV5XHDIW6XLBFWO84EqR+nTm6KakDRKB2TzZ/InUk7EPa89YELCxipn9zg3RVQpewxvxRWtGmBQ==","signature_status":"signed_v1","signed_at":"2026-05-20T02:05:41.640188Z","signed_message":"canonical_sha256_bytes"},"source_id":"2603.05910","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:11805b2d5bbcf3c9f90404f2f80da0a1671510e8d6d3d2e86a538431a35051a2","sha256:35e839682a012419622d745dcbe3d3bd7e5319be4a8fc9a254075d2271a76fdd"],"state_sha256":"277579460627270e82f5df96a8a27535292149e4034e68ca56866b6decf8bde4"}