{"api_version":"codesota.v1","data":{"current":{"created":1792329548,"proposals":[{"evidence":[{"body":"Qwen/Qwen3.5-0.8B completed 10042 examples (full split). Metric: accuracy_char_normalized. This records execution, not an independent validation or a global ranking.","id":"run:a4745c75b1996487651c72a8","occurred":1792253486,"title":"Measurement completed: HellaSwag (full validation, character-normalized accuracy)","url":"/runs/a4745c75b1996487651c72a8"},{"body":"Design version 1 applied bounded typography, table-spacing or desktop-width settings. Live settings are public; browser visual QA is not implied.","id":"design:1","occurred":1792328991,"title":"Readability settings updated","url":"/operations/design"}],"fact_ids":["run:a4745c75b1996487651c72a8","design:1"],"id":"benchmark-value-decision-view","next_step":"Test a summary view: top separations per benchmark, cost per separating example, benchmark and model-pair filters, capped first page with progressive disclosure. Measure time-to-conclusion for picking a cheaper benchmark that separates two named small models.","status":"proposal","title":"Make /benchmark-value answer a decision, not list 1645 rows","uncertainty":"Findings come from structural HTML analysis only; no rendered-browser or user study evidence yet.","why":"The page serves 1645 rows and 1.28 MB in 12.2 s with no visible filters, so pilots and full protocols blur together and a conclusion requires scrolling hundreds of rows."},{"evidence":[{"body":"Qwen/Qwen3.5-0.8B completed 3270 examples (full split). Metric: accuracy. This records execution, not an independent validation or a global ranking.","id":"run:130e0f8e238158a96c8145fe","occurred":1792256594,"title":"Measurement completed: BoolQ (full validation, raw likelihood accuracy)","url":"/runs/130e0f8e238158a96c8145fe"},{"body":"Publishes page-usability findings as an open improvement backlog. A recorded completed cycle demonstrates execution, not the correctness of every recommendation.","id":"agent:usability-20261018131724-410ebc","occurred":1792329478,"title":"Usability agent: first recorded completed cycle","url":"/api/agents/cycles/usability-20261018131724-410ebc"}],"fact_ids":["run:130e0f8e238158a96c8145fe","agent:usability-20261018131724-410ebc"],"id":"scope-visible-protocol-marking","next_step":"Test per-row scope badges, headings that name the pilot scope, and grouping rows by scope on home. Verify by asking a reader which figures come from full splits and checking they cite only pilot-marked rows.","status":"proposal","title":"Mark every number with its scope: pilot, full split, or cohort","uncertainty":"Grouping may already exist below the excerpt; verify the rendered page before changing anything.","why":"Pilot counts of 32 examples sit beside full-protocol rows with no per-row marker, and home rows mix scopes and outlier interest scales, so a skimmer may treat a 4-of-32 signal as a stable benchmark property."}],"summary":"Start from a concrete task: pick a benchmark and a model pair, see cost versus discrimination with pilot scope marked, and leave with one actionable conclusion. Prioritize the weak /benchmark-value path before adding anything new.","title":"Help people choose models and benchmarks with clear evidence","version":1},"policy":"Vision items are proposals, not measured results, commitments, or completed work. Evidence links are supplied by the server.","versions":[{"created":1792329548,"title":"Help people choose models and benchmarks with clear evidence","version":1}]}}