Onion Peeler

Browse the dark web, peel onions, and use AI to validate good and bad sites. Onion Peeler is a browser extension that transforms .onion links into accessible previews via the .pe domain. It helps users explore the dark web more safely and responsibly by integrating automatic content classification and moderation. 🌐 Smart .onion to .pe rendering Automatically appends .pe to .onion links so they can be previewed through a specialized gateway. Supports "Onion-Location" header integration and shows when a .onion.pe site is available. 🔒 Built-in local blocklist Checks links against a local SHA-256 hash list to block banned or harmful sites instantly, along with reasons for their ban, privately on your own device. 🧠 LLM-powered analysis Uses a backend system of cooperating large language models to analyze accessed onion sites. They summarize and categorize sites to assess safety and legality. 🚫 Dynamic link filtering If a site is flagged for promoting illegal content, its link is marked in red and becomes unclickable—ensuring users don’t inadvertently access harmful services. 📚 Ethical Indexing & Moderation By using Onion Peeler, you agree to the Terms of Service (https://onion.pe/tos.html) and Privacy Policy (https://onion.pe/pp.html). Onion URLs accessed via the .pe proxy may be logged and analyzed using AI to improve a public search index. This helps classify and suppress access to dangerous or illegal content. ⚠️ Disclaimer Onion Peeler does not host or promote dark web content. It simply provides a safer way to preview .onion links using the .pe proxy system and automated analysis. Accessing .onion links may expose users to material that is illegal or disturbing. This extension includes safeguards to block or label such links, but use is at your own risk. By installing this extension, you acknowledge that you are responsible for your browsing behavior and agree to comply with all applicable laws. 🚨 Abuse Reporting Found something harmful or wrongly classified? Please report abuse to [email protected]. We welcome SHA256 hashes of offending links or descriptions of the issue.