Skip to main content
Version: 0.2.3

Quickstart

Get Sercha up and running in 5 minutes. This guide walks you through installing Sercha, indexing your first documents, and running your first search.

Installation

brew tap custodia-labs/sercha
brew install sercha
First Run on macOS

macOS may block the binary on first run. If you see "killed", run:

xattr -d com.apple.quarantine $(which sercha)

Verify Installation

sercha version

Step 1: Add Your First Source

Let's index a local folder. This is the simplest way to get started.

sercha source add filesystem -c path=~/Documents

You should see:

Source created: <source-id>
Type: filesystem
Name: ~/Documents

Step 2: Sync Documents

Now sync the source to index your documents:

sercha sync

Sercha will scan the folder, normalise documents (PDF, Markdown, plain text, etc.), and build the search index.

Syncing source: <source-id>
Processed: 127 documents
Indexed: 127 documents
Duration: 3.2s
Sync a Specific Source

To sync only one source, pass its ID:

sercha sync <source-id>

Search your indexed documents:

sercha search "meeting notes"

Results show matching documents with relevance scores:

Found 5 results:

1. ~/Documents/work/meeting-notes-2024-01.md
Score: 0.92
...discussed Q1 planning and project timelines...

2. ~/Documents/projects/standup-notes.txt
Score: 0.78
...weekly meeting notes for the team...

Step 4: Launch the TUI

For an interactive experience, launch the Terminal UI:

sercha tui

The TUI provides:

  • Real-time search as you type
  • Keyboard navigation (j/k or arrows)
  • Document preview and actions
  • Source management
Keyboard Shortcuts
KeyAction
?Show help
qQuit
EnterSelect/execute
EscGo back
j/kNavigate up/down

Next Steps

Add More Sources

sercha source add github --token ghp_your_token -c content_types=files,issues
Getting a GitHub Token

Create a Personal Access Token at github.com/settings/tokens with repo scope.

See Supported Connectors for all options.

Sercha supports semantic search and LLM-assisted query expansion. Configure AI providers with the settings wizard:

sercha settings wizard

Cloud-based, requires API key from platform.openai.com.

  1. Run sercha settings wizard
  2. Select OpenAI as embedding and/or LLM provider
  3. Enter your API key when prompted

See AI Models for details on hybrid and AI-assisted search.

Learn More