scoring
Hosted MCP for denial, prior auth, reimbursement, workflow validation, batch scoring, and feedback.
我該用這個嗎
品質與安全性
發現項目（2）
- LOWTool 'score_workflow' description lacks action verb在 score_workflow 中
- LOWTool 'score_batch' description lacks action verb在 score_batch 中
根據工具定義與協定合規性的自動化分析。
上下文成本
這是每次將伺服器的工具載入模型上下文時所消耗的約略 token 數量。數量越高，可用於其他工作的注意力就越少。
安裝
一鍵安裝
將以下內容加入你的 `claude_desktop_config.json` 檔案：
{
"mcpServers": {
"scoring": {
"url": "https://mcp.sentinelsignal.io/mcp"
}
}
}遠端端點
https://mcp.sentinelsignal.io/mcpstreamable-http它能做什麼
工具清單
工具（8）
🟢list_workflows
List supported workflows and current model versions via GET /v1/workflows
輸入結構描述
{
"type": "object",
"properties": {}
}🟢get_workflow_schema(workflow)
Fetch required fields, optional fields, enums, and an example payload for a workflow via GET /v1/workflows/{workflow}/schema
輸入結構描述
{
"type": "object",
"properties": {
"workflow": {
"type": "string",
"description": "Workflow identifier such as healthcare.denial, healthcare.prior_auth, or healthcare.reimbursement."
}
},
"required": [
"workflow"
]
}🟢validate_workflow_payload(workflow, payload)
Validate and normalize a workflow payload without consuming a scoring call via POST /v1/workflows/{workflow}/validate
輸入結構描述
{
"type": "object",
"properties": {
"workflow": {
"type": "string",
"description": "Workflow identifier such as healthcare.denial, healthcare.prior_auth, or healthcare.reimbursement."
},
"payload": {
"type": "object",
"description": "Workflow payload to validate before scoring."
}
},
"required": [
"workflow",
"payload"
]
}🟢score_workflow(workflow, payload, options)
Score Sentinel healthcare workflows via POST /v1/score
輸入結構描述
{
"type": "object",
"properties": {
"workflow": {
"type": "string",
"description": "Workflow identifier such as healthcare.denial, healthcare.prior_auth, or healthcare.reimbursement."
},
"payload": {
"type": "object",
"description": "Structured workflow payload matching Sentinel /v1/score for the selected workflow."
},
"options": {
"type": [
"object",
"null"
],
"description": "Optional scoring flags such as explanation or threshold controls."
}
},
"required": [
"workflow",
"payload"
]
}🟢score_batch(items, continue_on_error)
Score up to 25 workflow items sequentially in one request via POST /v1/score/batch
輸入結構描述
{
"type": "object",
"properties": {
"items": {
"type": "array",
"description": "List of scoring items, each containing workflow, payload, and optional options."
},
"continue_on_error": {
"type": "boolean",
"description": "When true, later items continue even if an earlier item fails."
}
},
"required": [
"items"
]
}🟢get_limits
Retrieve plan limits, trial quota, and upgrade state via GET /v1/limits
輸入結構描述
{
"type": "object",
"properties": {}
}🟢get_usage(month)
Retrieve monthly scoring call usage via GET /v1/usage
輸入結構描述
{
"type": "object",
"properties": {
"month": {
"type": "string",
"description": "Optional billing month in YYYY-MM format."
}
}
}🟡submit_feedback(feedback)
Submit structured claims outcome feedback via POST /v1/feedback
輸入結構描述
{
"type": "object",
"properties": {
"feedback": {
"type": "object",
"description": "Structured feedback payload for a prior Sentinel scoring event."
}
},
"required": [
"feedback"
]
}建議的提示詞
get_workflow_schemaget_workflow_schemalist_workflowslist_workflowslist_workflowsget_workflow_schema社群
證據