aboutsummaryrefslogtreecommitdiff
path: root/test-suite/notes/scoring.md
diff options
context:
space:
mode:
authorTianren Dong <tianrendong.dtr@gmail.com>2026-05-13 07:29:41 -0400
committerTianren Dong <tianrendong.dtr@gmail.com>2026-05-13 07:29:41 -0400
commit4c54b05dc15b81e449dc8ac53da8bae536fa2798 (patch)
treef81c5e627dcc4dd81d2fe2fd3b2b584e72a73eca /test-suite/notes/scoring.md
parent5951b3fb04aaf53b875afe265a57a944d6fce641 (diff)
pi-chrome: snapshot pre-migration state
Diffstat (limited to 'test-suite/notes/scoring.md')
-rw-r--r--test-suite/notes/scoring.md36
1 files changed, 36 insertions, 0 deletions
diff --git a/test-suite/notes/scoring.md b/test-suite/notes/scoring.md
new file mode 100644
index 0000000..970a706
--- /dev/null
+++ b/test-suite/notes/scoring.md
@@ -0,0 +1,36 @@
+# Benchmark scoring
+
+## Unit challenges (`manifest.json`)
+
+Score actual verdict against `expected[mode]`, not raw PASS count.
+
+- `PASS` / `FAIL` expected: exact match required.
+- `CONDITIONAL`: exclude from headline score unless environment prerequisite is declared present.
+- `SKIP` / `WARN`: report separately.
+
+Recommended flake policy:
+
+- Run each non-destructive challenge up to 2 retries.
+- Take best verdict only for known-flaky timing/scroll tests.
+- Always keep all reasons/events in detailed report.
+
+Difficulty weighting:
+
+- L1 = 1
+- L2 = 2
+- L3 = 3
+
+Report both unweighted and weighted score. Also report per-category score.
+
+## Long-horizon tasks (`task-manifest.json`)
+
+Task score is deterministic grader PASS / not PASS. Record:
+
+- action count
+- wall time
+- tools used
+- observation mode (`snapshot`, screenshot, accessibility, evaluate)
+- whether direct state mutation/evaluate was allowed
+- final grader reason
+
+Do not use LLM judge for hermetic tasks unless grader cannot express the target.