创思大模型安全
创思大模型安全MCP是面向大语言模型的内容安全防护系统，通过实时风险识别与拦截，保障大模型应用的安全、合规与可信。其核心能力包括输入/输出内容双重安全检测、多维度风险识别（合规/伦理/安全）、低延迟实时防护以及可定制策略引擎。
MCP 服务配置
复制以下 JSON 到 OPClaw 或其他 MCP 客户端的配置文件中即可使用
{
"mcpServers": {
"chuangsiai": {
"name": "chuangsiai-mcp",
"type": "sse",
"url": "https://mcp.chuangsiai.com/sse"
}
}
}该服务需要配置环境变量：Authorization、StrategyKey
服务介绍
Chuangsi Large Model Security MCP
Official Website
Introduction
Chuangsi Large Model Security MCP (Model Content Protection) is a content security protection system designed for large language models. It ensures the security, compliance, and trustworthiness of model applications through real-time risk identification and interception.
Core Capabilities:
- Dual security checks on input/output content
- Multi-dimensional risk identification (compliance/ethics/security)
- Low-latency real-time protection
- Customizable policy engine
Quick Integration
Add the following service configuration in your configuration file (supports SSE protocol):
json
{
"mcpServers": {
"chuangsiai": {
"name": "chuangsiai-mcp",
"type": "sse", // Streaming transmission protocol
"baseUrl": "https://mcp.chuangsiai.com/sse",
"headers": {
"Authorization": "【Your API Key】",
"StrategyKey": "【Your Strategy Identifier】"
}
}
}
}
Credential Acquisition Steps
- Log in to the Chuangsi Security Console
- Go to "API Key Management" to create an API Key
- Obtain the strategy identifier from the "Policy Center"
- Replace the placeholders 【】in the configuration
Technical Features
- Real-time Risk Identification: Quickly identifies potential risks such as non-compliant content and ethical issues by analyzing input and output content in real time.
- Multi-dimensional Detection: Supports multiple risk detection dimensions, including but not limited to compliance, ethics, and security.
- Low Latency: Uses HTTPS transmission protocol to ensure low-latency real-time protection, meeting real-time requirements.
- Customizable Policy Engine: Provides a flexible policy engine that supports user-defined risk detection rules and protection strategies.
Use Cases
- Content Moderation: Real-time monitoring and moderation of input and output content to ensure it complies with laws, regulations, and core socialist values.
📬 Contact Us
For technical support, business cooperation, or API integration, please contact:
- Email: service@chuangsiai.com
- Official Website: https://chuangsiai.com
- Console: https://console.chuangsiai.com
