Test fixture
Bug fixes, edge conditions, and minimal patch accuracy.
The model receives the prompt (and optional system message). The run uses scorer rubric_json_metrics with the JSON configuration below. Pass/fail and partial credit are determined entirely by that scorer against the model output; no human grading.
Return JSON only with keys diagnosis, fix, tests. A memoized lookup returns cached misses as null and treats them as cache hits, so later populated values are never read. Identify the bug and propose a minimal JavaScript fix.
{
"metrics": {
"repro": {
"checks": [
{
"contains": [
"cached miss"
]
},
{
"contains": [
"null"
]
},
{
"contains": [
"later populated"
]
},
{
"contains": [
"cache hit"
]
}
]
},
"hidden": {
"checks": [
{
"contains": [
"hasOwn"
]
},
{
"contains": [
"Map.has"
]
},
{
"contains": [
"undefined sentinel"
]
}
]
},
"diagnose": {
"checks": [
{
"contains": [
"memoized lookup"
]
},
{
"contains": [
"truthy"
]
},
{
"contains": [
"miss"
]
}
]
}
}
}temperature
0
max_tokens
420
timeout (s)
120
type
scored
file
debug-nullish-cache-hit-v2.json