Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/s-hari-s/aether
A simple rag based plugin for obsidian
https://github.com/s-hari-s/aether
gemini-api obsidian-md obsidian-plugin rag retrieval-augmented-generation
Last synced: 7 days ago
JSON representation
A simple rag based plugin for obsidian
- Host: GitHub
- URL: https://github.com/s-hari-s/aether
- Owner: S-HARI-S
- Created: 2024-09-22T13:52:16.000Z (3 months ago)
- Default Branch: master
- Last Pushed: 2024-10-31T05:58:47.000Z (about 2 months ago)
- Last Synced: 2024-10-31T06:21:53.491Z (about 2 months ago)
- Topics: gemini-api, obsidian-md, obsidian-plugin, rag, retrieval-augmented-generation
- Language: TypeScript
- Homepage:
- Size: 13.7 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Obsidian RAG Plugin
A simple Retrieval-Augmented Generation (RAG) plugin for Obsidian that allows you to query your vault using Google's Gemini AI.
## Features
- Real-time monitoring of markdown files in your vault
- Automatic embedding generation and updates
- Semantic search using ChromaDB
- Conversation history support
- Markdown-aware text chunking
- File change detection (create, modify, delete)
## Setting Up
1. Clone the repository into your Obsidian vault's plugins directory:
```bash
cd YOUR_VAULT_PATH/.obsidian/plugins
git clone [repository-url] rag-plugin
```
2. Set up the Python environment:
```bash
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
```
3. Install required Python packages:
```bash
pip install flask chromadb google-generativeai python-dotenv watchdog
```
4. Configure the environment:
- Add your Google AI API key:
```
# Configure Google AI
genai.configure(api_key="")
```
- Update `ROOT_DIR` in the code to point to your Obsidian vault path
- ```
Here:
ROOT_DIR = r"C:\Users\SREEHARI\Documents\Obsidian Vault"
```
5. Start the Python backend server:
```bash
python new.py
```
6. Install Node.js dependencies and start the frontend:
```bash
npm install
npm run dev
```
7. Enable the plugin in Obsidian:
- Go to Settings → Community Plugins
- Enable the plugin named "sample pluggin"
## Quick glance at Technical Details
- **Database**: Uses ChromaDB for vector storage (in-memory mode)
- **Embedding Strategy**:
- Chunks markdown files by sections and paragraphs
- Maximum chunk size: 1000 characters
- Maintains document references and metadata
- **AI Model**: Uses Gemini 1.5 Flash with configured parameters:
- Temperature: 0.7
- Top-p: 0.95
- Top-k: 64
- Max output tokens: 1000
- **File Processing**:
- Monitors `.md` files only
- Ignores hidden directories
- Supports real-time updates
- **API Endpoint**:
- POST `/arraysum` for querying the knowledge base ( "arraysum" doesn't make sense, i will change this eventually )
- Maintains conversation history (last 5 exchanges)