{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:NP4IZKUYWJ2OF5ZXXR5S32MIKJ","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":"bafb25d41e092c73ba83da09e53c4d209e084068eab17522d44c574fa189c148","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.SE","submitted_at":"2026-06-17T21:36:09Z","title_canon_sha256":"de2cdc96eb966a9c6ec763f2f778b69ad41d48c1e11af27e5176c49918995c57"},"schema_version":"1.0","source":{"id":"2606.19613","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.19613","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"arxiv_version","alias_value":"2606.19613v1","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.19613","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"pith_short_12","alias_value":"NP4IZKUYWJ2O","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"pith_short_16","alias_value":"NP4IZKUYWJ2OF5ZX","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"pith_short_8","alias_value":"NP4IZKUY","created_at":"2026-06-19T16:12:30Z"}],"graph_snapshots":[{"event_id":"sha256:2703cd9a20d0118b9bc9eb59c3b9f70099f323b90f445f36765f1671b46139f0","target":"graph","created_at":"2026-06-19T16:12:30Z","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/2606.19613/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We introduce StaminaBench, a benchmark that measures the stamina of coding agents: how many consecutive interaction turns (change requests) they can handle before failing. Unlike the prevailing fraction-of-tasks-solved metric, this matches real vibe-coding where sessions run dozens or hundreds of turns. In StaminaBench, agents implement a REST API server and modify it across a tunable number of procedurally generated follow-up change requests - 100 in our experiments, resulting in codebases of up to 6,000 lines. Tests are generated fully programmatically without LLM involvement, ensuring repro","authors_text":"Shuo Yang, Stefano Soatto, Vlad Sobal, Wei Xia, Yuting Zhang","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.SE","submitted_at":"2026-06-17T21:36:09Z","title":"StaminaBench: Stress-Testing Coding Agents over 100 Interaction Turns"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.19613","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":"sha256:86a2c91b4991b5934748a5d7266c6d42a7c2b74232aee72d5e35b9a3862d4622","target":"record","created_at":"2026-06-19T16:12:30Z","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":"bafb25d41e092c73ba83da09e53c4d209e084068eab17522d44c574fa189c148","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.SE","submitted_at":"2026-06-17T21:36:09Z","title_canon_sha256":"de2cdc96eb966a9c6ec763f2f778b69ad41d48c1e11af27e5176c49918995c57"},"schema_version":"1.0","source":{"id":"2606.19613","kind":"arxiv","version":1}},"canonical_sha256":"6bf88caa98b274e2f737bc7b2de9885255c27ad8ec8c44dad560babb187f0508","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6bf88caa98b274e2f737bc7b2de9885255c27ad8ec8c44dad560babb187f0508","first_computed_at":"2026-06-19T16:12:30.203782Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-19T16:12:30.203782Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"lXTp7RcOT8wXVlVs/1+9OLVKBGj2x+xKnvLzJv0zMKLbJPLwYtTotRJZv71BIsVAe8hN55rQRGEeJPSPjcCxCA==","signature_status":"signed_v1","signed_at":"2026-06-19T16:12:30.204146Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.19613","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:86a2c91b4991b5934748a5d7266c6d42a7c2b74232aee72d5e35b9a3862d4622","sha256:2703cd9a20d0118b9bc9eb59c3b9f70099f323b90f445f36765f1671b46139f0"],"state_sha256":"9d03f59abec7de868df7929dbaf00041d84c1b615e57b269ec9451b2ce4a7735"}