About AdRadar
AdRadar is an MCP service for competitor ad research in AI agents, aimed at marketers and researchers. It searches public Meta and TikTok ads and analyzes creative hooks, angles, and marketing signals.
Key features
- 01Search the public Meta Ad Library
- 02View top TikTok auction ads
- 03Score ad copy across ten dimensions
- 04Return ad performance and creative signals
- 05Check remaining scan credits
Use cases
- 01Compare competitors' ads that are still running
- 02Research hooks and angles across ads
- 03Find top TikTok ads by country
- 04Evaluate ad copy before launch
MCP support
AdRadar provides an MCP service over Streamable HTTP at https://adradarmcp.cloud/mcp, with OAuth 2.1 or bearer-key authentication. MCP clients can call its ad search, evaluation, and credit-check tools.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://adradarmcp.cloud/mcpclaude mcp add --transport http egggy1998-adradar https://adradarmcp.cloud/mcp{"mcpServers":{"egggy1998-adradar":{"type":"http","url":"https://adradarmcp.cloud/mcp"}}}Open in Cursor{"servers":{"egggy1998-adradar":{"type":"http","url":"https://adradarmcp.cloud/mcp"}}}Open in VS Code{"mcpServers":{"egggy1998-adradar":{"type":"http","url":"https://adradarmcp.cloud/mcp"}}}{"mcpServers":{"egggy1998-adradar":{"serverUrl":"https://adradarmcp.cloud/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
https://adradarmcp.cloud/mcpAuth requiredTools
Details
From the project READMEAdRadar MCP
Competitor ad research inside your AI agent. Ask Claude, ChatGPT or Cursor which Meta and TikTok ads your competitors keep running, and get them scored and explained.
- Website: https://adradarmcp.cloud/en/?utm_source=github&utm_medium=directory&utm_campaign=launch_w1
- MCP endpoint (Streamable HTTP):
https://adradarmcp.cloud/mcp - Auth: OAuth 2.1 (claude.ai, ChatGPT) or
Authorization: Bearer <your API key>(or?key=<your API key>for clients that only accept a URL) - Pricing: $0.05 per scan, $1 for 20 scans to try. Credits never expire. Failed or empty scans are refunded automatically.
This repository holds the public listing (README +
server.json) for directories and the MCP Registry. AdRadar is a hosted service; the server code is not open source.
Tools
| Tool | What it does | Cost |
|---|---|---|
ad_radar_spy | Search the public Meta Ad Library by keyword and country. Returns active ads with run time, creative reuse, hook type, angle, audience, CTA strength, urgency, price, guarantee and social proof signals. Default country: US. | 1 credit |
ad_radar_tiktok_top | Top TikTok auction ads from TikTok Creative Center for a country: CTR, likes, budget tier, objective, video URL. | 1 credit |
ad_radar_evaluate | Score any ad copy on a 10-dimension marketing matrix, no scraping. | 1 credit |
ad_radar_balance | Remaining credits. | Free |
Setup
Claude (claude.ai / desktop) and ChatGPT: add a custom connector with URL https://adradarmcp.cloud/mcp and sign in when prompted (OAuth 2.1). You can also use https://adradarmcp.cloud/mcp?key=<your API key>.
Claude Code:
claude mcp add --transport http adradar https://adradarmcp.cloud/mcp --header "Authorization: Bearer <your API key>"
Cursor / Windsurf / any MCP client (mcp.json):
{
"mcpServers": {
"adradar": {
"url": "https://adradarmcp.cloud/mcp",
"headers": { "Authorization": "Bearer <your API key>" }
}
}
}
Example prompts
- "Find the longest-running protein powder ads in the US and group them by angle."
- "Compare the hooks of my top 3 competitors' active ads and suggest 5 angles nobody is using."
- "Score this ad copy and tell me what to fix before I launch it."
- "What are the top TikTok ads in the UK this month, and what objective are they optimizing for?"
Data and privacy
AdRadar only reads ads that Meta and TikTok publish in their public ad libraries. Privacy policy: https://adradarmcp.cloud/en/privacy
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.