Hister is an open-source search engine designed to index and search content from web pages you visit and files you keep locally. According to its GitHub repository, it enables full-text search of visited pages and documents through a web interface, terminal, or AI assistant connected via MCP (Model Context Protocol).
The tool operates on a local-first model, with no mandatory cloud service or telemetry. Users download a binary for their platform (Linux, macOS, or Windows), start a local server, and install a browser extension for Firefox or Chrome. Once enabled, the extension automatically indexes newly visited pages. Hister then allows users to search across indexed content through multiple interfaces.
Key features include full-text indexing of actual page contents rather than just titles and URLs, support for advanced queries using field filters, phrases, wildcards, and negation, and optional semantic search via configurable embeddings endpoints. The tool can also import existing browser history, crawl websites, and index local directories.
Hister supports multi-user setups on shared servers while keeping each user’s documents and search results separate. By default, the browser extension sends indexed page content only to the Hister server configured by the user, with the exception of favicon downloads. All documents and search indexes are stored locally on that server.
For users who opt into optional semantic search, document text is sent to embeddings endpoints they choose. The project emphasizes reviewing privacy settings before enabling remote integrations.
Installation methods include downloading precompiled binaries, using package managers like Homebrew, Docker, or Nix, or building from source using Go 1.26, npm, and a C compiler. The project is licensed under AGPLv3 or any later version and welcomes contributions via its GitHub issue tracker and community channels including IRCNet and Discord.
Key facts
- Hister is a locally-hosted search engine for indexing web pages and files with no required cloud service
- It supports full-text search through a web interface, terminal, or AI assistant via MCP
- Browser extensions for Firefox and Chrome automatically index visited pages
- The tool includes advanced query features like field filters, wildcards, and optional semantic search
- All indexed content and search results remain on the user’s own Hister server by default
- Multi-user support allows separate document and result management per user on shared servers