trace_synthesis/tasks/497.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

37 lines
984 B
JSON

{
"sites": [
"shopping_admin"
],
"task_id": 497,
"require_login": true,
"storage_state": "./.auth/shopping_admin_state.json",
"start_url": "http://localhost:28083/admin",
"geolocation": null,
"intent_template": "Update order #{{order}} with the {{service}} tracking number {{tracking}}",
"instantiation_dict": {
"tracking": "24353446464",
"order": "307",
"service": "DHL"
},
"intent": "Update order #307 with the DHL tracking number 24353446464",
"require_reset": false,
"eval": {
"eval_types": [
"program_html"
],
"reference_answers": null,
"reference_url": "",
"program_html": [
{
"url": "http://localhost:28083/admin/sales/order/commentsHistory/order_id/307/active_tab/order_shipments/",
"locator": "",
"required_contents": {
"must_include": [
"Tracking number 24353446464 for DHL assigned"
]
}
}
]
},
"intent_template_id": 284
}