historySearchAI

historySearchAI is a Chrome extension that combines two complementary functions: advanced browser cleaning and AI-assisted history search.

The extension first makes it easy to quickly clean the main categories of browser data through a simple and direct interface. Users can remove browsing history, downloads, cache, cookies, saved passwords, and form data, then automate some of these actions on a chosen schedule. The goal is to provide a practical tool for keeping the browser cleaner, better organized, and easier to manage on a daily basis.

The core feature of historySearchAI is the AI Search area, a dedicated conversational window designed to explore Chrome history more effectively. Instead of manually browsing long and often hard-to-read lists, users can ask natural-language questions to find what they previously visited more easily. For example, they can search for videos watched recently, find pages related to a specific topic, explore recent browsing activity by subject, or filter results by date, title, domain, or URL fragment.

The extension is designed to remain practical and grounded. It first relies on the data actually available in Chrome history, then uses AI to help formulate, sort, filter, and work with the results. Users can then open the matching links directly or selectively remove specific URLs that were found. This approach makes browsing history significantly more usable than simply reviewing Chrome’s native history page.

historySearchAI supports multiple AI providers: OpenAI, Anthropic, Google Gemini, as well as a local model accessible through a compatible HTTP endpoint URL. This allows users to choose between a cloud service and local infrastructure depending on their needs, preferences, or privacy requirements. No AI model is included in the extension: access to AI features depends on an external API key or a user-configured local endpoint.

The extension is also designed for international use. It supports multiple languages, allows users to manually select the interface language in preferences, and adapts model instructions according to the selected language and expected date format. Settings also make it possible to test an API key or local endpoint and choose which model to use for the selected provider.