Ocolta Review Summaries
Read categorical signals, confidence and indicator counts from your recoverable Ocolta reviews.
我该使用它吗
质量与安全性
基于对工具定义和协议合规性的自动分析。
上下文开销
这是每次将服务器的工具加载到模型上下文窗口时所消耗的大致 token 数。数值越高，可用于其他任务的注意力就越少。
安装
一键安装
将以下内容添加到你的 `claude_desktop_config.json` 文件中：
{
"mcpServers": {
"ocolta-review-summaries": {
"url": "https://app.ocolta.com/mcp"
}
}
}远程端点
https://app.ocolta.com/mcpstreamable-http它能做什么
工具清单
工具（3）
🟢get_connected_purchase
Identify the existing Ocolta purchase connected for this conversation. Returns available review count and supported report-summary access; does not expose payment details or account identifiers.
输入模式
{
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"label": {
"type": "string"
},
"synthetic": {
"type": "boolean"
},
"availableReviews": {
"type": "integer",
"minimum": 0
},
"access": {
"type": "string",
"const": "Categorical summaries from recoverable completed reviews"
}
},
"required": [
"label",
"synthetic",
"availableReviews",
"access"
],
"additionalProperties": false
}🟢list_recent_reviews
List up to ten completed, recoverable Deep Reviews from the existing purchase you connected. Use when you ask which review summaries are still available within the 24-hour recovery period. No new review or credit spend is started.
输入模式
{
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"reviews": {
"type": "array",
"maxItems": 10,
"items": {
"type": "object",
"properties": {
"review_id": {
"type": "string",
"pattern": "^ent_[a-f0-9]{32}(?:c[0-9]{2,3})?$"
}
},
"required": [
"review_id"
],
"additionalProperties": false
}
},
"recoveryWindow": {
"type": "string",
"const": "24 hours after completion"
}
},
"required": [
"reviews",
"recoveryWindow"
],
"additionalProperties": false
}🟢get_review_summary(review_id)
Retrieve categorical document-review status, review priority, confidence and fraud/AI indicator counts for one recoverable review from the connected purchase. Provides signals for human follow-up without source documents, personal fields or free-text excerpts. This does not prove fraud, authenticity, intent, identity or AI generation. No new credit is spent.
输入模式
{
"type": "object",
"properties": {
"review_id": {
"type": "string",
"pattern": "^ent_[a-f0-9]{32}(?:c[0-9]{2,3})?$"
}
},
"required": [
"review_id"
],
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"completed",
"unable_to_determine",
"refused"
]
},
"reviewPriority": {
"type": "string",
"enum": [
"low",
"moderate",
"high",
"unable_to_determine"
]
},
"confidence": {
"type": "string",
"enum": [
"low",
"medium",
"high",
"unable_to_determine"
]
},
"fraudAssessment": {
"type": "string",
"enum": [
"no_supported_indicators",
"indicators_present",
"unable_to_determine"
]
},
"aiGenerationAssessment": {
"type": "string",
"enum": [
"no_supported_indicators",
"indicators_present",
"unable_to_determine"
]
},
"fraudIndicatorCount": {
"type": "integer",
"minimum": 0
},
"aiGenerationIndicatorCount": {
"type": "integer",
"minimum": 0
},
"limitationCount": {
"type": "integer",
"minimum": 0
},
"disclaimer": {
"type": "string",
"const": "This review identifies document-level signals for human follow-up. It does not prove fraud, authenticity, intent, identity, account ownership, or AI generation."
}
},
"required": [
"status",
"reviewPriority",
"confidence",
"fraudAssessment",
"aiGenerationAssessment",
"fraudIndicatorCount",
"aiGenerationIndicatorCount",
"limitationCount",
"disclaimer"
],
"additionalProperties": false
}社区
证据