Hister: A Private, Full-Content Search Index Under Your Control

Hister transforms the digital trail of your web browsing and local file storage into a private, comprehensive search index that you own and manage entirely.

Rather than relying on fragile bookmarks or vague filenames, Hister ensures that the knowledge you've already discovered remains accessible and searchable.

🚀 Core Value Proposition

Hister allows you to ~~retrace your steps manually~~ jump straight to the exact idea you remember. By storing a full-content preview alongside the index, you can verify the context of a result without ever leaving your search page.

"A private index makes the things you already found useful again."

Key Pillars of Hister

Full Content Preservation: It doesn't just index titles; it saves the actual text.
Absolute Privacy: No telemetry, no "phoning home," and no mandatory cloud dependencies.
Self-Hosted Sovereignty: Run it on your own hardware or a private server.
Open Source: Distributed as free software under the AGPLv3 license.

⚙️ How the System Operates

Hister functions as a seamless loop of data collection, processing, and retrieval.

1. Collection

You can populate your index through several channels:

Web Pages: Automatically save sites as you visit them via the browser extension.
Local Files: Monitor specific folders for changes.
History: Import existing browser history.
Crawling: Target specific websites for deep indexing.

2. Indexing

The Hister server extracts the essential data and creates a searchable map. This process can be fine-tuned using:

Skip rules: To ignore irrelevant content.
Priority rules: To boost the importance of specific sources.

3. Retrieval

Access your data through multiple interfaces. Whether you prefer a GUI or a command line, your data is available. For those using AI, Hister supports the Model Context Protocol (MCP), allowing AI assistants to retrieve your private data.

🔍 Precision Searching

Hister goes beyond simple keyword matching. It utilizes a sophisticated query language to narrow down results.

Search Capabilities:

Feature	Description	Example
Field Filters	Limit search to specific metadata	`title: "Privacy"`
Quoted Phrases	Search for exact strings	`"digital sovereignty"`
Wildcards/Negation	Broaden or exclude terms	`search* -spam`
Query Aliases	Create shortcuts for complex queries	`@notes`

The mathematical logic of the ranking can be thought of as: $\text{Relevance} = \sum (\text{Keyword Frequency} \times \text{Priority Weight}) - \text{Penalty}_{\text{skip}}$

🛡️ The Privacy Model

Hister is designed to be a "black box" that only you have the key to.

Zero Telemetry: The server never reports your searches or indexed content.
Local-First: A complete setup can exist on a single offline machine.
Controlled Routing: Clients only communicate with the server you specify.
Auditable: The AGPLv3 source code is open for public inspection.

🛠️ Deployment & Technical Specs

Getting Hister running is flexible, ranging from a simple binary to a full-scale database setup.

Installation Options

Quickstart: Single binary for local use.
Database: Use SQLite for light use or PostgreSQL for heavy loads.
Containerized: Deploy via Docker or Nix.
Multi-user: Shared servers with user-scoped access.

Developer Interface

You can interact with your index programmatically via the HTTP API:

# Example: Querying the Hister API via curl
curl -X GET "http://localhost:8080/search?q=privacy+notes" \
     -H "Authorization: Bearer YOUR_TOKEN"

Advanced Capabilities

Semantic Search: Optional integration with an embeddings endpoint for concept-based searching.
Version Tracking: Keep track of how indexed content changes over time.
Language Awareness: Specialized indexes for different languages.
Custom Extractors: Handle structured data from various website formats.