Overview

Sercha is a privacy-first local search engine that indexes your personal data and makes it instantly searchable. All data stays on your machine.

Why Sercha?

Feature	Description
Privacy-first	Your data never leaves your machine
Local-first	Works offline, no internet required
Extensible	Plugin system for data sources and file formats
Hybrid search	Full-text (Xapian) + vector similarity (HNSWlib)

How It Works

1. Connectors fetch raw data from your sources
2. Normalisers transform content into searchable text
3. Index stores everything locally for instant search

Supported Connectors

Connector	Description	Authentication
Filesystem	Local files and directories	None
GitHub	Repositories, issues, PRs, wikis	PAT or OAuth
Google Drive	Documents, spreadsheets, files	OAuth
Gmail	Email messages and threads	OAuth
Google Calendar	Calendar events	OAuth

See Supported Connectors for the full list.

Search Modes

Sercha supports multiple search modes depending on your AI provider configuration:

Mode	Description	Requirements
Text Only	Fast keyword search using Xapian	None (default)
Hybrid	Keyword + semantic vector search	Embedding provider
LLM Assisted	Query expansion with AI	LLM provider
Full	All features combined	Both providers

See AI Models for configuration details.

Interfaces

Sercha provides multiple ways to interact:

Interface	Description	Status
CLI	Command-line for scripting and quick searches	Available
TUI	Interactive terminal UI with keyboard navigation	Available
MCP	Model Context Protocol server for AI tools	Available

Architecture

Sercha uses Hexagonal Architecture (Ports & Adapters) to enable:

• Adding data sources without changing core code
• Adding file formats without changing core code
• Swapping search backends without affecting connectors
• Parallel development across teams

See the Architecture Overview for details.

Next Steps

• Quickstart - Get up and running in 5 minutes
• Basic Configuration - Configure your setup