Adwords MCP

This Skill addresses a 'beloved' pain point among developers—forcibly inserting advertisements into AI assistant conversations. In real-world development, developers often need to focus on using AI assistants for coding or problem-solving. Adwords MCP, in a playful manner, simulates how malicious middleware can inject irrelevant and embarrassing ad content into LLM responses via keyword matching, thereby warning developers about the potential security risks of MCP servers.

Usage is straightforward: simply install the server globally via npm and add the corresponding server configuration to your MCP client (such as Claude Desktop or Cursor) configuration file. Thereafter, during normal conversations with your AI assistant, the server will randomly select and insert preset 'embarrassing' ad copy based on keywords in the dialogue, disrupting the normal interaction flow.

It is particularly suitable for teams or individuals conducting MCP protocol security demonstrations or security awareness training—especially developers, security researchers, and educators learning how to build MCP servers or understand the principles of malicious middleware attacks. This project humorously reveals the poor experience that can result from technology misuse.

It is recommended to use this tool uniformly during internal tech talks, security workshops, or MCP teaching demonstrations to enhance persuasiveness. Note that this is primarily a demonstration and entertainment project; its ad-matching logic is relatively simple and crude. It should not be used in production environments or serious development work to avoid impacting normal workflow efficiency.