Skip to content

Actions: michalstrnadel/scorekeeper

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
127 workflow runs
127 workflow runs

Filter by Workflow

⌥ + click/return to exclude

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
data(results): d8cx phase 2 partial — opus-4-8 holds 3/3, bare ORR 0/2 ci #120: Commit b847998 pushed by michalstrnadel
36s main
main
36s
data(results): Gemini drip — first same-evening discordant pair (s03:… ci #119: Commit b2b5544 pushed by michalstrnadel
5m 30s main
main
5m 30s
data(results): URR smoke — deny holds until the ledger records the re… ci #118: Commit 49f9c74 pushed by michalstrnadel
29s main
main
29s
docs(assets): social preview card — the real board as the repo's face ci #117: Commit 905fed5 pushed by michalstrnadel
31s main
main
31s
data(results): barge-seed stability — 0/9 on re-run; the barge is a d… ci #116: Commit df98c22 pushed by michalstrnadel
27s main
main
27s
data(results): powered set 2 — stronger elicitation, gpt-5.4-mini sta… ci #115: Commit 4d0fd15 pushed by michalstrnadel
36s main
main
36s
data(results): first powered seed set — 10 overreach pairs on gpt-5.4… ci #114: Commit 9661853 pushed by michalstrnadel
31s main
main
31s
chore: release 0.4.0 — cross-model bench, scorekeeper board release #7: Commit 902c635 pushed by michalstrnadel
1m 2s v0.4.0
v0.4.0
1m 2s
chore: release 0.4.0 — cross-model bench, scorekeeper board ci #113: Commit 902c635 pushed by michalstrnadel
32s main
main
32s
docs: What the board looks like — carried by the board GIF ci #112: Commit 587935e pushed by michalstrnadel
30s main
main
30s
docs: surface the cross-model result everywhere it was owed ci #111: Commit 4cd5ec1 pushed by michalstrnadel
27s main
main
27s
docs(results): silent placebo — board presence does nothing, the dige… ci #110: Commit dbe78d2 pushed by michalstrnadel
28s main
main
28s
docs(results): loop replication — two backends, three seeds; bare 3/4… ci #109: Commit 4220f5b pushed by michalstrnadel
29s main
main
29s
docs(results): LOOP-SMOKE — first cross-model evidence, the barge rep… ci #108: Commit bb798b3 pushed by michalstrnadel
25s main
main
25s
data(results): arming study raw output (F20 evidence) ci #107: Commit 67cbe63 pushed by michalstrnadel
30s main
main
30s
docs(results): fold F20/F21 into the head — both-on owed, arming meas… ci #106: Commit c7145f9 pushed by michalstrnadel
28s main
main
28s
docs(results): F20 — arming reliability is ~94% over valid extractions ci #105: Commit 8fef953 pushed by michalstrnadel
26s main
main
26s
fix(bench): reclassify drops degraded runs instead of re-scoring them ci #104: Commit d961334 pushed by michalstrnadel
32s main
main
32s
docs: the strong-model both-on cell is owed, not held (F21) ci #103: Commit 0631df2 pushed by michalstrnadel
33s main
main
33s
docs(results): F21 — the seeded both-on cell is dropped (credits died… ci #102: Commit 2077886 pushed by michalstrnadel
32s main
main
32s
fix(bench): exhausted usage credits are a lost turn, not a behavior ci #101: Commit 9354f3c pushed by michalstrnadel
33s main
main
33s
docs(results): record the serial-subscription rule and Fable cell costs ci #100: Commit 1a490db pushed by michalstrnadel
28s main
main
28s
feat(bench): arming_study — measure how often extraction arms the wall ci #99: Commit f8f5c59 pushed by michalstrnadel
27s main
main
27s
test(bench): cover the harness wiring the F19 fixes depend on ci #98: Commit d890f5f pushed by michalstrnadel
29s main
main
29s
chore(bench): relock harness after the 0.3.2 core bump ci #97: Commit 7aed964 pushed by michalstrnadel
28s main
main
28s