trace_synthesis/tasks/494.json
yuyr a84d51a101 1. 增加r1生成综合策略代码和输出;
2. 增加tasks;
3. 增加analysis部分,对策略进行归纳分类,然后进行评测。
2025-04-17 17:40:15 +08:00

34 lines
1.1 KiB
JSON

{
"sites": [
"shopping_admin"
],
"task_id": 494,
"require_login": true,
"storage_state": "./.auth/shopping_admin_state.json",
"start_url": "http://localhost:28083/admin",
"geolocation": null,
"intent_template": "Notify {{name}} in their most recent pending order with message \"{{message}}\"",
"instantiation_dict": {
"name": "Alex Thomas",
"message": "Yo, your order will be shipped soon!"
},
"intent": "Notify Alex Thomas in their most recent pending order with message \"Yo, your order will be shipped soon!\"",
"require_reset": false,
"eval": {
"eval_types": [
"program_html"
],
"reference_answers": null,
"reference_url": "",
"program_html": [
{
"url": "http://localhost:28083/admin/sales/order/view/order_id/304/",
"locator": "document.querySelector(\"#order_history_block\").querySelector(\".note-list\").firstElementChild.querySelector(\".note-list-comment\").outerText",
"required_contents": {
"exact_match": "Yo, your order will be shipped soon!"
}
}
]
},
"intent_template_id": 280
}