{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:5K53UDX6QNRFDJ4TL5KM7Z2Y7I","short_pith_number":"pith:5K53UDX6","schema_version":"1.0","canonical_sha256":"eabbba0efe836251a7935f54cfe758fa1f5a55765a3886d4bcb68b179aeda599","source":{"kind":"arxiv","id":"2603.05910","version":2},"attestation_state":"computed","paper":{"title":"The World Won't Stay Still: Programmable Evolution for Agent Benchmarks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","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","submitted_at":"2026-03-06T04:56:18Z","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"},"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":"2603.05910","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-03-06T04:56:18Z","cross_cats_sorted":[],"title_canon_sha256":"47ca4f3bf33cd76be5962b42017385a4893e6a8fb7c84cec9bae736adce9fac2","abstract_canon_sha256":"325a977d04776de69d549e0771ca519b851b00e2a9da34503d0511e87dd3534f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T02:05:41.640188Z","signature_b64":"OWsnNWBRL1rfV5XHDIW6XLBFWO84EqR+nTm6KakDRKB2TzZ/InUk7EPa89YELCxipn9zg3RVQpewxvxRWtGmBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eabbba0efe836251a7935f54cfe758fa1f5a55765a3886d4bcb68b179aeda599","last_reissued_at":"2026-05-20T02:05:41.639197Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T02:05:41.639197Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The World Won't Stay Still: Programmable Evolution for Agent Benchmarks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","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","submitted_at":"2026-03-06T04:56:18Z","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"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.05910","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/2603.05910/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":"2603.05910","created_at":"2026-05-20T02:05:41.639364+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.05910v2","created_at":"2026-05-20T02:05:41.639364+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.05910","created_at":"2026-05-20T02:05:41.639364+00:00"},{"alias_kind":"pith_short_12","alias_value":"5K53UDX6QNRF","created_at":"2026-05-20T02:05:41.639364+00:00"},{"alias_kind":"pith_short_16","alias_value":"5K53UDX6QNRFDJ4T","created_at":"2026-05-20T02:05:41.639364+00:00"},{"alias_kind":"pith_short_8","alias_value":"5K53UDX6","created_at":"2026-05-20T02:05:41.639364+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/5K53UDX6QNRFDJ4TL5KM7Z2Y7I","json":"https://pith.science/pith/5K53UDX6QNRFDJ4TL5KM7Z2Y7I.json","graph_json":"https://pith.science/api/pith-number/5K53UDX6QNRFDJ4TL5KM7Z2Y7I/graph.json","events_json":"https://pith.science/api/pith-number/5K53UDX6QNRFDJ4TL5KM7Z2Y7I/events.json","paper":"https://pith.science/paper/5K53UDX6"},"agent_actions":{"view_html":"https://pith.science/pith/5K53UDX6QNRFDJ4TL5KM7Z2Y7I","download_json":"https://pith.science/pith/5K53UDX6QNRFDJ4TL5KM7Z2Y7I.json","view_paper":"https://pith.science/paper/5K53UDX6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.05910&json=true","fetch_graph":"https://pith.science/api/pith-number/5K53UDX6QNRFDJ4TL5KM7Z2Y7I/graph.json","fetch_events":"https://pith.science/api/pith-number/5K53UDX6QNRFDJ4TL5KM7Z2Y7I/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5K53UDX6QNRFDJ4TL5KM7Z2Y7I/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5K53UDX6QNRFDJ4TL5KM7Z2Y7I/action/storage_attestation","attest_author":"https://pith.science/pith/5K53UDX6QNRFDJ4TL5KM7Z2Y7I/action/author_attestation","sign_citation":"https://pith.science/pith/5K53UDX6QNRFDJ4TL5KM7Z2Y7I/action/citation_signature","submit_replication":"https://pith.science/pith/5K53UDX6QNRFDJ4TL5KM7Z2Y7I/action/replication_record"}},"created_at":"2026-05-20T02:05:41.639364+00:00","updated_at":"2026-05-20T02:05:41.639364+00:00"}