Tool tree
Review benchmark steps, inputs, orchestrator prompts, and verification evidence.
Implementation
gradeNot graded
gradeNot graded
filter_benchmark_scenariosAction
Filter the visible benchmark scenarios by text and review status. Does not submit or modify reviews.
View tool JSON
{
"name": "filter_benchmark_scenarios",
"kind": "act",
"impl": "imperative",
"description": "Filter the visible benchmark scenarios by text and review status. Does not submit or modify reviews.",
"inputSchema": {
"type": "object",
"properties": {
"query": {
"type": "string"
},
"status": {
"type": "string",
"enum": [
"all",
"unreviewed",
"correct",
"revision",
"unsure"
]
}
},
"additionalProperties": false
},
"page": "/spectral_agent_bench/"
}