Skip to content

MCP-Development-Guide.md

Latest commit

 

History

History
43 lines (29 loc) · 1.85 KB

MCP-Development-Guide.md

File metadata and controls

43 lines (29 loc) · 1.85 KB

MCP Development Guide

How to build MCP (Model Context Protocol) servers for advertising platforms.

What Is MCP?

MCP is a protocol that lets AI tools (Claude, Cursor, Gemini) call external APIs through a standardized interface. An MCP server for Google Ads, for example, lets Claude pull live campaign data, run reports, and make changes — all within a conversation.

Why Build Advertising MCP Servers?

Every platform has an API. But connecting an AI tool to that API requires:

  1. Authentication handling
  2. Request formatting
  3. Response parsing
  4. Error handling
  5. Rate limiting

An MCP server wraps all of this into tools that AI can call naturally.

Getting Started

  1. Choose a platform — Check mcp-servers/registry.yaml for what's available
  2. Start from a template — Python template or Node template
  3. Implement the standard tools — See SPEC.md
  4. Use core/ for auth — core/auth/<platform>.py handles credentials
  5. Test with Claude Code — Add to your claude_desktop_config.json
  6. Submit a PR — Add to the registry and share with the community

Standard Tool Interface

Every advertising MCP server should implement at minimum:

  • list_campaigns — List campaigns with status and metrics
  • get_campaign — Detailed campaign data
  • get_metrics — Performance metrics for date ranges
  • list_audiences — Available audience segments
  • get_budget — Budget and pacing status

See SPEC.md for full specifications.

Reference Implementation

The google-ads-mcp server is the reference implementation. Study its architecture for patterns you can apply to other platforms.