What is Hister?
Hister is a private, self-hosted search engine that turns the pages you visit and the files you keep into a full-content search index that you control. Instead of relying on bookmarks or filenames, it indexes the actual text of the documents you choose, so you can find things again by what they say rather than where they live.
How it works
Hister follows a simple loop:
- Collect — a browser extension saves newly visited pages, while file watchers, history imports, and a website crawler add other sources.
- Index — content extractors pull out the parts that matter and index their full text on the server you configure.
- Find — search from the web interface, terminal, command line, HTTP API, or let an AI assistant retrieve results through MCP.
Key features
- Full-content search — look beyond titles and URLs to the words inside every indexed document.
- Precise queries — field filters, quoted phrases, wildcards, negation, date ranges, priorities, and custom aliases.
- Stored previews — open a clean, readable preview of the saved content beside your results.
- Multiple interfaces — web UI, terminal client, CLI, HTTP API, and MCP server all share the same index.
- Flexible deployment — a single binary can run locally; shared servers support user-scoped access with SQLite or PostgreSQL, plus Docker and Nix.
- Index rules — skip and priority rules control indexing and ranking, with optional version tracking.
- Optional semantic search — sends text only to the embeddings endpoint you configure.
Privacy by design
The index, stored page content, and rules stay on the Hister server you set up. There is no telemetry, no mandatory cloud service, and no external requests by default. The source is public and licensed under AGPLv3, so the software is auditable.
Who it is for
Hister suits anyone who wants durable, private recall of their own reading and documents — researchers, developers, writers, and privacy-conscious users who prefer to own their search infrastructure rather than rent it.
