{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:MXWJYBYRETJ465A6GXA3FOZJ6D","short_pith_number":"pith:MXWJYBYR","schema_version":"1.0","canonical_sha256":"65ec9c071124d3cf741e35c1b2bb29f0e5a52bbb2b381a7e938ff11afa3aee92","source":{"kind":"arxiv","id":"2605.22535","version":1},"attestation_state":"computed","paper":{"title":"TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chao Peng, Earl T. Barr, Federica Sarro, Han Li, He Ye, Jiarui Hu, Mark Harman, Pengyu Zou, Peter O'Hearn, Xingyu Jiang, Zhaoyang Chu","submitted_at":"2026-05-21T14:24:43Z","abstract_excerpt":"We introduce TerminalWorld, a scalable data engine that automatically reverse-engineers high-fidelity evaluation tasks from \"in-the-wild\" terminal recordings. Processing 80,870 terminal recordings, the engine yields a full benchmark of 1,530 validated tasks, spanning 18 real-world categories, ranging from short everyday operations to workflows exceeding 50 steps, and covering 1,280 unique commands. From these, we curate a Verified subset of 200 representative, manually reviewed tasks. Comprehensive benchmarking on TerminalWorld-Verified across eight frontier models and six agents reveals that "},"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":"2605.22535","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-21T14:24:43Z","cross_cats_sorted":[],"title_canon_sha256":"2671f032a60f8234d74c71f1801a3d82577003307d9a8e4fecc9b8288934cc3f","abstract_canon_sha256":"9dfdff4b80c9984d8f4474d112ea88319c034ba251e86646b9999c6dcd0a3a9f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-22T01:04:56.105662Z","signature_b64":"AAMasywqKt5nrE/Q9xX+25VDD6niuNBsXjX9mMpewdhiemGit25nbOgLCxLq4D3JYO0YZcgOPrq3Qcbn2m9XCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"65ec9c071124d3cf741e35c1b2bb29f0e5a52bbb2b381a7e938ff11afa3aee92","last_reissued_at":"2026-05-22T01:04:56.104569Z","signature_status":"signed_v1","first_computed_at":"2026-05-22T01:04:56.104569Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chao Peng, Earl T. Barr, Federica Sarro, Han Li, He Ye, Jiarui Hu, Mark Harman, Pengyu Zou, Peter O'Hearn, Xingyu Jiang, Zhaoyang Chu","submitted_at":"2026-05-21T14:24:43Z","abstract_excerpt":"We introduce TerminalWorld, a scalable data engine that automatically reverse-engineers high-fidelity evaluation tasks from \"in-the-wild\" terminal recordings. Processing 80,870 terminal recordings, the engine yields a full benchmark of 1,530 validated tasks, spanning 18 real-world categories, ranging from short everyday operations to workflows exceeding 50 steps, and covering 1,280 unique commands. From these, we curate a Verified subset of 200 representative, manually reviewed tasks. Comprehensive benchmarking on TerminalWorld-Verified across eight frontier models and six agents reveals that "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.22535","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/2605.22535/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":"2605.22535","created_at":"2026-05-22T01:04:56.104730+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.22535v1","created_at":"2026-05-22T01:04:56.104730+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.22535","created_at":"2026-05-22T01:04:56.104730+00:00"},{"alias_kind":"pith_short_12","alias_value":"MXWJYBYRETJ4","created_at":"2026-05-22T01:04:56.104730+00:00"},{"alias_kind":"pith_short_16","alias_value":"MXWJYBYRETJ465A6","created_at":"2026-05-22T01:04:56.104730+00:00"},{"alias_kind":"pith_short_8","alias_value":"MXWJYBYR","created_at":"2026-05-22T01:04:56.104730+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/MXWJYBYRETJ465A6GXA3FOZJ6D","json":"https://pith.science/pith/MXWJYBYRETJ465A6GXA3FOZJ6D.json","graph_json":"https://pith.science/api/pith-number/MXWJYBYRETJ465A6GXA3FOZJ6D/graph.json","events_json":"https://pith.science/api/pith-number/MXWJYBYRETJ465A6GXA3FOZJ6D/events.json","paper":"https://pith.science/paper/MXWJYBYR"},"agent_actions":{"view_html":"https://pith.science/pith/MXWJYBYRETJ465A6GXA3FOZJ6D","download_json":"https://pith.science/pith/MXWJYBYRETJ465A6GXA3FOZJ6D.json","view_paper":"https://pith.science/paper/MXWJYBYR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.22535&json=true","fetch_graph":"https://pith.science/api/pith-number/MXWJYBYRETJ465A6GXA3FOZJ6D/graph.json","fetch_events":"https://pith.science/api/pith-number/MXWJYBYRETJ465A6GXA3FOZJ6D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MXWJYBYRETJ465A6GXA3FOZJ6D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MXWJYBYRETJ465A6GXA3FOZJ6D/action/storage_attestation","attest_author":"https://pith.science/pith/MXWJYBYRETJ465A6GXA3FOZJ6D/action/author_attestation","sign_citation":"https://pith.science/pith/MXWJYBYRETJ465A6GXA3FOZJ6D/action/citation_signature","submit_replication":"https://pith.science/pith/MXWJYBYRETJ465A6GXA3FOZJ6D/action/replication_record"}},"created_at":"2026-05-22T01:04:56.104730+00:00","updated_at":"2026-05-22T01:04:56.104730+00:00"}