ai-compliance-monitor

Regulatory intelligence for AI agents across jurisdictions

我该使用它吗

质量与安全性

A
描述质量
95%
模式完整度
97%
命名质量
95%
投毒风险
100%
权限匹配度
100%
协议合规性
100%

基于对工具定义和协议合规性的自动分析。

上下文开销

~583token 数(工具定义)
~1.2 KB典型响应大小
对注意力的影响极小(占 128k 上下文窗口的 0.46%)

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": {
    "ai-compliance-monitor": {
      "url": "https://ai-compliance-monitor.sgdata.workers.dev/mcp"
    }
  }
}

远程端点

https://ai-compliance-monitor.sgdata.workers.dev/mcpstreamable-http

它能做什么

工具清单

工具(4)

🟢 只读🟡 写入🔴 删除⚪ 未知
🟢check_obligations(use_case, jurisdiction_code)

Given an AI agent use case and optional jurisdiction, returns all applicable regulatory obligations with evidence requirements, enforcement dates, and penalty information. Use this to understand what compliance requirements apply to a specific type of AI agent.

输入模式

{
  "type": "object",
  "properties": {
    "use_case": {
      "type": "string",
      "description": "The AI agent use case to check. Examples: hiring_screening, credit_scoring, customer_service, content_moderation, medical_triage, autonomous_coding, financial_trading"
    },
    "jurisdiction_code": {
      "type": "string",
      "description": "Optional jurisdiction code to filter by. Examples: EU, SG, US-CO. Omit to get all jurisdictions."
    }
  },
  "required": [
    "use_case"
  ]
}
🟢get_regulation_articles(regulation_id, category)

Returns structured, machine-readable regulation details including specific articles, obligation classifications, risk levels, and evidence requirements. Use this to understand the detailed requirements of a specific regulation.

输入模式

{
  "type": "object",
  "properties": {
    "regulation_id": {
      "type": "string",
      "description": "The regulation identifier. Options: eu-ai-act, sg-imda-agentic, co-ai-act"
    },
    "category": {
      "type": "string",
      "description": "Optional filter by obligation category. Options: transparency, record_keeping, human_oversight, risk_assessment, incident_reporting, data_governance, monitoring, audit"
    }
  },
  "required": [
    "regulation_id"
  ]
}
🟢check_deadline(jurisdiction_code, include_passed, months_ahead)

Returns upcoming and recent compliance deadlines with enforcement dates, penalty information, and related obligations. Use this to understand time-sensitive compliance requirements.

输入模式

{
  "type": "object",
  "properties": {
    "jurisdiction_code": {
      "type": "string",
      "description": "Optional jurisdiction code to filter by. Examples: EU, SG, US-CO"
    },
    "include_passed": {
      "type": "boolean",
      "description": "Whether to include deadlines that have already passed. Default: false"
    },
    "months_ahead": {
      "type": "number",
      "description": "Number of months ahead to look for upcoming deadlines. Default: 12"
    }
  },
  "required": []
}
⚪compare_jurisdictions(category, jurisdictions)

Side-by-side comparison of regulatory obligations across jurisdictions for a given compliance category. Shows equivalent, overlapping, stricter, and weaker mappings between regulations. Use this to understand how requirements differ across borders.

输入模式

{
  "type": "object",
  "properties": {
    "category": {
      "type": "string",
      "description": "The obligation category to compare. Options: transparency, record_keeping, human_oversight, risk_assessment, incident_reporting, data_governance, monitoring, audit"
    },
    "jurisdictions": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "Optional list of jurisdiction codes to compare. Default: all available. Examples: [\"EU\", \"SG\", \"US-CO\"]"
    }
  },
  "required": [
    "category"
  ]
}

社区

评价此服务器

证据

最近观测

已验证未记录版本4 个工具
已验证未记录版本4 个工具
已验证未记录版本4 个工具