JasonLovesDoggo/caddy-defender

What it solves

It prevents unwanted traffic and protects website owners from having their content scraped by AI services to train models. It specifically addresses the problem of AI bots polluting or harvesting data by providing tools to block these services or feed them misleading information.

How it works

This is a middleware plugin for the Caddy web server. It filters incoming requests based on the client's IP address. It comes with a built-in list of IP ranges for major AI services (like OpenAI, DeepSeek, and GitHub Copilot) and cloud providers. When a match is found, the plugin can trigger different responses: it can block the request (403 Forbidden), drop the connection, redirect the user, or use "garbage" and "tarpit" modes to send back useless data or slow down the connection to waste the bot's resources.

Who it’s for

Website administrators and server owners who use Caddy and want to prevent AI crawlers from scraping their data for training purposes.

Highlights

• Predefined AI IP Ranges: Includes ready-to-use IP lists for services like OpenAI, Mistral, and DeepSeek.
• Anti-Scraping Tactics: Offers specialized responders like "garbage" and "tarpit" to actively pollute AI training data or stall bots.
• Flexible Filtering: Supports custom CIDR ranges and integration with other Caddy plugins like caddy-ratelimit.
• Multiple Response Options: Can return custom messages, 403 errors, or permanent redirects.