Database · MODEL CONTEXT PROTOCOL
gregce adwords mcp
A cringe-worthy MCP server that serves ads to developers in Cursor, Claude, and other clients
WHAT IT CONNECTS
gregce adwords mcp 为 AI Agent 提供什么
- 🔍 **Naive Keyword Extraction**: Using simplistic string matching to ensure maximum ad interruption
- 🎯 **Random Ad Selection**: Chooses ads based on detected keywords or just randomly if no keywords match
- 💥 **Cringe Ad Injection**: Multiple strategies for embedding ads in responses
- 🔄 **Multiple Transport Options**: Primarily STDIO-based with HTTP/SSE support
- 📝 **Resource Templates**: Access ad templates through MCP resources (optional)
- 🛠️ **Configurable Options**: Customize behavior through command-line flags or programmatic API
- ⚡ **Tool Aliases**: Short aliases for all tools to make invocation easier
- Clone the repository:
- Install dependencies:
- Build the project:
- Link it globally
- (oPTIONAL) Run the server in development mode:
- **Tool-based API**: Exposes several tools through the Model Context Protocol:
- `get_completion`: Returns a completion with embedded advertisements
- `analyze_code`: Performs fake code analysis with embedded advertisements
- `developer_tip`: Provides development tips with embedded advertisements
- Aliases: `gc`, `ac`, and `tip` as shorter alternatives
- **Keyword-based Ad Targeting**: Matches keywords in user queries to serve targeted ads
- **Multiple Ad Injection Strategies**: Uses various techniques to inject ads into responses
SECURITY
MCP 收录不等于安全审核
MCP 服务器可能获得模型上下文、凭据、本地文件或调用外部系统的权限。连接 Agent 前,请检查代码、环境变量、网络行为、软件包来源与维护状态。