### Reasoning with Ollama

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Example configuration for enabling reasoning with Ollama, specifying models for reasoning and fast processing.

```javascript
OLLAMA_MODEL=<REASONING_MODEL>
OLLAMA_FAST_MODEL=non reasoning model
REASONING=true
FAST_MODEL_REASONING=false
```

--------------------------------

### Deploy Xyne with Docker Compose

Source: https://github.com/xynehq/xyne/blob/main/docs/quickstart.mdx

Starts the Xyne application using Docker Compose. Requires Docker and Docker Compose to be installed and running. This command builds and runs the necessary containers, making the app available on port 3001.

```shell
docker-compose -f deployment/docker-compose.yml up
```

--------------------------------

### Reasoning with Together AI

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Example configuration for enabling reasoning with Together AI, specifying models for reasoning and fast processing.

```javascript
TOGETHER_API_KEY=<YOUR_TOGETHER_API_KEY>
TOGETHER_MODEL=deepseek-ai/DeepSeek-R1
TOGETHER_FAST_MODEL=deepseek-ai/DeepSeek-V3
REASONING=true
FAST_MODEL_REASONING=false
```

--------------------------------

### Clone Xyne Repository (Git)

Source: https://github.com/xynehq/xyne/blob/main/docs/quickstart.mdx

Clones the Xyne repository from GitHub. Requires Git to be installed. This command fetches all necessary project files to your local machine.

```shell
git clone https://github.com/xynehq/xyne.git
```

--------------------------------

### Reasoning with Fireworks AI

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Example configuration for enabling reasoning with Fireworks AI, specifying models for reasoning and fast processing.

```javascript
FIREWORKS_API_KEY=<YOUR_FIREWORKS_API_KEY>
FIREWORKS_MODEL=accounts/fireworks/models/deepseek-r1
FIREWORKS_FAST_MODEL=accounts/fireworks/models/deepseek-v3
REASONING=true
FAST_MODEL_REASONING=false
```

--------------------------------

### Start Promtail using PM2

Source: https://github.com/xynehq/xyne/blob/main/observability/docs/observability.md

Command to start the Promtail service using PM2, specifying the configuration file. This ensures Promtail runs persistently and can be managed by PM2.

```bash
pm2 start <full-path-to-installed-promtail> -- \
  -config.file=<full-path-to-xyne>/deployment/promtail-config.yaml
```

--------------------------------

### Build Xyne Container with Docker Compose

Source: https://github.com/xynehq/xyne/blob/main/docs/quickstart.mdx

Command to build and start the Xyne application container using a Docker Compose file. This is a crucial step for deploying or running the application.

```bash
docker-compose -f deployment/docker-compose.yml up
```

--------------------------------

### Install pm2-prom-module

Source: https://github.com/xynehq/xyne/blob/main/observability/docs/observability.md

Installs the pm2-prom-module using the PM2 command-line interface. This module enables PM2 to export Prometheus-compatible metrics.

```sh
pm2 install pm2-prom-module
```

--------------------------------

### Run Docs Locally

Source: https://github.com/xynehq/xyne/blob/main/docs/README.md

This command initiates the local development server for the project's documentation. It requires Bunx to be installed and uses Mintlify to serve the documentation locally.

```shell
bunx mintlify dev
```

--------------------------------

### Setup Virtual Environment and Install pytrec_eval

Source: https://github.com/xynehq/xyne/blob/main/server/eval/README.md

Sets up a Python virtual environment using 'uv' and installs the 'pytrec_eval' package. This is a prerequisite for running the final evaluation metrics calculation.

```shell
uv venv
source .venv/bin/activate
uv pip install pytrec_eval
```

--------------------------------

### Configure Ollama Model

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Sets up the Ollama model name for local LLM usage in Xyne. Requires OLLAMA_MODEL. Ensure the specified model is downloaded and Ollama is running.

```javascript
OLLAMA_MODEL=<YOUR_OLLAMA_MODEL_NAME>
```

--------------------------------

### Configure and Run Client

Source: https://github.com/xynehq/xyne/blob/main/docs/contribution/devmode.mdx

Sets up the client-side environment by creating a `.env` file with the API base URL and then installs dependencies and starts the development server using Bun. The client application will be available on port 5173.

```javascript
cd frontend
touch .env
# Add VITE_API_BASE_URL=http://localhost:3000 and VITE_WS_BASE_URL=ws://localhost:3000 to .env
bun install
bun run dev
```

--------------------------------

### Run Metrics Dashboard with Docker Compose

Source: https://github.com/xynehq/xyne/blob/main/observability/docs/observability.md

Starts the metrics dashboard services using Docker Compose. This command assumes the deployment file is located in the root folder.

```sh
docker-compose -f deployment/docker-compose.metrics.yml up
```

--------------------------------

### Enable Google APIs

Source: https://github.com/xynehq/xyne/blob/main/docs/authentication/service-accounts.mdx

Enables specific Google APIs required for application integration. This involves navigating the Google Cloud Console, searching for APIs, and enabling them.

```APIDOC
API: Google Drive API
Purpose: Access and manage Google Drive files.
Status: Enable

API: Google Calendar API
Purpose: Access and manage Google Calendar events.
Status: Enable

API: Google Slides API
Purpose: Create and manage Google Slides presentations.
Status: Enable

API: Google Docs API
Purpose: Create and manage Google Docs documents.
Status: Enable

API: Google Sheets API
Purpose: Create and manage Google Sheets spreadsheets.
Status: Enable

API: Gmail API
Purpose: Send and receive emails via Gmail.
Status: Enable

API: Contacts API
Purpose: Access and manage Google Contacts.
Status: Enable

API: People API
Purpose: Access and manage user profile information.
Status: Enable

API: Admin SDK
Purpose: Manage Google Workspace administration tasks.
Status: Enable
```

--------------------------------

### Configure OpenAI API Key

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Sets up the OpenAI API key for Xyne's chat feature. Requires OPENAI_API_KEY.

```javascript
OPENAI_API_KEY=<YOUR_OPENAI_API_KEY>
```

--------------------------------

### Enable Reasoning Configuration

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Enables reasoning functionality for supported AI providers by setting the REASONING environment variable to 'true'.

```javascript
REASONING=true
```

--------------------------------

### Initialize and Run Server

Source: https://github.com/xynehq/xyne/blob/main/docs/contribution/devmode.mdx

Navigates to the server directory, executes the initialization script to set up server dependencies, and then starts the development server using Bun. The server will listen on port 3000.

```javascript
cd server
./init-script.sh
bun run dev
```

--------------------------------

### Install LibreOffice (macOS, Ubuntu, RHEL)

Source: https://github.com/xynehq/xyne/blob/main/server/integrations/dataSource/README.md

Provides installation commands for LibreOffice, a required external tool for converting Office documents to PDF for processing.

```bash
# macOS
brew install --cask libreoffice

# Ubuntu/Debian
sudo apt-get install libreoffice

# RHEL/CentOS
sudo yum install libreoffice
```

--------------------------------

### Configure Together AI Credentials and Models

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Sets up Together AI API key and model names for Xyne's chat feature. Requires TOGETHER_API_KEY, TOGETHER_MODEL, and TOGETHER_FAST_MODEL.

```javascript
TOGETHER_API_KEY=<YOUR_TOGETHER_API_KEY>
TOGETHER_MODEL=<YOUR_TOGETHER_MODEL>
TOGETHER_FAST_MODEL=<YOUR_TOGETHER_FAST_MODEL>
```

--------------------------------

### Configure Gemini API Key and Model

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Sets up Google Gemini API key and model name for Xyne's chat feature. Requires GEMINI_API_KEY and GEMINI_MODEL.

```javascript
GEMINI_API_KEY=<YOUR_GEMINI_API_KEY>
GEMINI_MODEL=<YOUR_GEMINI_MODEL_NAME>
```

--------------------------------

### Stop Xyne Docker Compose

Source: https://github.com/xynehq/xyne/blob/main/docs/quickstart.mdx

Stops and removes the containers created by Docker Compose. Requires Docker and Docker Compose to be installed. This command cleans up the running Xyne services.

```shell
docker-compose -f deployment/docker-compose.yml down
```

--------------------------------

### Configure Base URL for OpenAI or Together AI

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Allows specifying a custom base URL for OpenAI or Together AI services when using their respective API keys.

```javascript
OPENAI_API_KEY=<YOUR_OPEN_API_KEY>
    ### OR
TOGETHER_API_KEY=<YOUR_TOGETHER_API_KEY>
BASE_URL=<YOUR_BASE_URL>
```

--------------------------------

### Google Cloud Project & OAuth Consent Screen Setup

Source: https://github.com/xynehq/xyne/blob/main/docs/authentication/get-credentials.mdx

Steps to create a Google Cloud project and configure the OAuth consent screen. This involves setting app information, user support email, audience type, contact information, and defining necessary scopes for authentication.

```APIDOC
GoogleCloudSetup:
  Steps:
    - Create Project in Google Cloud Console
    - Configure OAuth Consent Screen:
        - App Information:
            - Name: [Your App Name] (e.g., Xyne)
            - User Support Email: [Your Support Email] (e.g., your-email@gmail.com)
        - Audience:
            - Type: Internal (or External for personal emails)
            - Test Users: [Optional, for External audience]
        - Contact Information:
            - Developer Contact Email: [Your Contact Email]
        - Scopes:
            - Add: /auth/userinfo.email, /auth/userinfo.profile, openid
            - Leave out others initially unless required.
        - User Data Policy Acceptance: Accept Google's User Data Policy.
```

--------------------------------

### Configure Fireworks AI Credentials and Models

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Sets up Fireworks AI API key and model names for Xyne's chat feature. Requires FIREWORKS_API_KEY, FIREWORKS_MODEL, and FIREWORKS_FAST_MODEL.

```javascript
FIREWORKS_API_KEY=<YOUR_FIREWORKS_API_KEY>
FIREWORKS_MODEL=<YOUR_FIREWORKS_MODEL>
FIREWORKS_FAST_MODEL=<YOUR_FIREWORKS_FAST_MODEL>
```

--------------------------------

### Install ImageMagick (macOS, Ubuntu, RHEL)

Source: https://github.com/xynehq/xyne/blob/main/server/integrations/dataSource/README.md

Provides installation commands for ImageMagick, a required external tool for converting image files for processing.

```bash
# macOS
brew install imagemagick

# Ubuntu/Debian
sudo apt-get install imagemagick

# RHEL/CentOS
sudo yum install ImageMagick
```

--------------------------------

### Start Docker Containers

Source: https://github.com/xynehq/xyne/blob/main/docs/contribution/devmode.mdx

Starts Vespa and PostgreSQL containers using Docker Compose for local development. This command ensures that the necessary backend services are running.

```bash
docker-compose -f deployment/docker-compose.dev.yml up
```

--------------------------------

### Development Environment Configuration

Source: https://github.com/xynehq/xyne/blob/main/server/integrations/dataSource/README.md

Example environment variables for configuring the data source file upload system during development, setting file size limits, conversion timeouts, and temporary directory paths.

```bash
export DATASOURCE_MAX_FILE_SIZE_MB=5
export DATASOURCE_CONVERSION_TIMEOUT_MS=15000
export DATASOURCE_TEMP_DIR=/tmp/datasource
```

--------------------------------

### Configure Custom AI Endpoints

Source: https://github.com/xynehq/xyne/blob/main/docs/quickstart.mdx

Sets the base URL for custom AI providers like OpenAI or Together AI. Ensure this variable points to your specific endpoint.

```shell
BASE_URL=<YOUR_BASE_URL>
HOST=http://localhost:3001
```

--------------------------------

### Execute Search Quality Evaluation Script

Source: https://github.com/xynehq/xyne/blob/main/server/scripts/README.md

Demonstrates running the `evaluateSearchQuality.ts` script using `bun`. Examples cover basic execution, enabling detailed failure analysis (`DEBUG_POOR_RANKINGS=true`), and customizing parameters like `EVALUATION_STRATEGY` and `NUM_SAMPLES` via environment variables.

```bash
# Basic usage (calculates metrics only):
EVALUATION_USER_EMAIL="user@example.com" bun run server/scripts/evaluateSearchQuality.ts

# Run with failure analysis enabled:
EVALUATION_USER_EMAIL="user@example.com" \
DEBUG_POOR_RANKINGS=true \
bun run server/scripts/evaluateSearchQuality.ts

# Example with a specific strategy and more samples:
EVALUATION_USER_EMAIL="user@example.com" \
EVALUATION_STRATEGY="BodyPhrase" \
NUM_SAMPLES=200 \
DEBUG_POOR_RANKINGS=true \
bun run server/scripts/evaluateSearchQuality.ts
```

--------------------------------

### AI Provider Key Selection Order

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Defines the precedence order for AI provider keys if multiple are set in environment variables. The system uses the first available key in this sequence.

```javascript
AWS → OpenAI → Ollama → Together AI → Fireworks AI → Gemini
```

--------------------------------

### Configure Environment Variables (.env.default)

Source: https://github.com/xynehq/xyne/blob/main/docs/quickstart.mdx

Defines environment variables for Xyne configuration, including authentication keys, API keys for various AI providers, and database/Vespa hostnames. These variables are crucial for the application's functionality and security.

```env
ENCRYPTION_KEY=<YOUR_ENCRYPTION_KEY>
SERVICE_ACCOUNT_ENCRYPTION_KEY=<YOUR_SERVICE_ACCOUNT_ENCRYPTION_KEY>

GOOGLE_CLIENT_ID=<YOUR_GOOGLE_CLIENT_ID>
GOOGLE_CLIENT_SECRET=<YOUR_GOOGLE_CLIENT_SECRET>
GOOGLE_REDIRECT_URI=http://localhost:3000/v1/auth/callback
GOOGLE_PROD_REDIRECT_URI=http://localhost:3001/v1/auth/callback

JWT_SECRET=<YOUR_JWT_SECRET>

DATABASE_HOST=xyne-db
VESPA_HOST=vespa

## If using AWS Bedrock
AWS_ACCESS_KEY=<YOUR_AWS_ACCESS_KEY>
AWS_SECRET_KEY=<YOUR_AWS_SECRET_ACCESS>
AWS_REGION=<YOUR_AWS_REGION>
  
## OR [ If using Open AI ]
OPENAI_API_KEY=<YOUR_OPEN_API_KEY>

## OR [ If using Ollama ] 
OLLAMA_MODEL=<YOUR_OLLAMA_MODEL_NAME> 

## OR [ If using Together AI ] 
TOGETHER_API_KEY=<YOUR_TOGETHER_API_KEY>
TOGETHER_MODEL=<YOUR_TOGETHER_MODEL>
TOGETHER_FAST_MODEL=<YOUR_TOGETHER_FAST_MODEL>

## OR [ If using Fireworks AI ] 
FIREWORKS_API_KEY=<YOUR_FIREWORKS_API_KEY>
FIREWORKS_MODEL=<YOUR_FIREWORKS_MODEL>
FIREWORKS_FAST_MODEL=<YOUR_FIREWORKS_FAST_MODEL>

## OR [If using Google AI]
GEMINI_API_KEY=<YOUR_GEMINI_API_KEY> 
GEMINI_MODEL=<YOUR_GEMINI_MODEL_NAME>
```

--------------------------------

### Example Test Query Structures

Source: https://github.com/xynehq/xyne/blob/main/eval-data/README.md

Illustrates various types of test queries, including factual, temporal, personal, and data search. Each example demonstrates the required 'input' and 'expected' fields within a JSON object.

```JSON
{
  "input": "what is my email",
  "expected": "user@example.com"
}
```

```JSON
{
  "input": "when was my last meeting",
  "expected": "Yesterday at 3pm with Marketing team"
}
```

```JSON
{
  "input": "what's my job title",
  "expected": "Senior Developer"
}
```

```JSON
{
  "input": "find emails about project alpha",
  "expected": "Found 3 emails from last week about project alpha"
}
```

--------------------------------

### Configure AWS Credentials

Source: https://github.com/xynehq/xyne/blob/main/docs/apikeys/apikeys.mdx

Sets up AWS API keys and region for Xyne's chat feature. Requires AWS_ACCESS_KEY, AWS_SECRET_KEY, and optionally AWS_REGION. The AWS_REGION defaults to 'us-west-2' if not provided.

```javascript
AWS_ACCESS_KEY=<YOUR_AWS_ACCESS_KEY>
AWS_SECRET_KEY=<YOUR_AWS_SECRET_KEY>
AWS_REGION=<YOUR_AWS_REGION>
```

--------------------------------

### Integrate ESLint React Plugin

Source: https://github.com/xynehq/xyne/blob/main/frontend/README.md

This snippet demonstrates how to integrate the `eslint-plugin-react` into your ESLint configuration. It includes setting the React version and enabling recommended rules from the plugin.

```javascript
// eslint.config.js
import react from 'eslint-plugin-react'

export default tseslint.config({
  // Set the react version
  settings: { react: { version: '18.3' } },
  plugins: {
    // Add the react plugin
    react,
  },
  rules: {
    // other rules...
    // Enable its recommended rules
    ...react.configs.recommended.rules,
    ...react.configs['jsx-runtime'].rules,
  },
})
```

--------------------------------

### Grant Permissions for Docker Loki Directory

Source: https://github.com/xynehq/xyne/blob/main/observability/docs/observability.md

Command to grant recursive read/write/execute permissions to the deployment/loki directory, often needed to resolve permission issues when running Docker containers.

```bash
sudo chmod -R 777 deployment/loki
```

--------------------------------

### Running Xyne with Docker Compose

Source: https://github.com/xynehq/xyne/blob/main/README.md

Xyne can be deployed locally or in the cloud using Docker Compose. This command facilitates easy setup for running the Xyne application.

```bash
docker-compose up
```

--------------------------------

### Download NanoBeir Datasets

Source: https://github.com/xynehq/xyne/blob/main/server/eval/README.md

Installs the 'datasets' package using 'uv pip' and then runs the 'getNanoBeir.py' script to download all NanoBeir datasets. This is an alternative method for obtaining evaluation datasets.

```shell
uv pip install datasets
python getNanoBeir.py
```

--------------------------------

### Example Error Handling

Source: https://github.com/xynehq/xyne/blob/main/server/integrations/dataSource/README.md

Illustrates how to catch and handle custom data source errors, specifically checking if an error is a data source error and accessing its machine-readable code and user-friendly message.

```typescript
try {
  const result = await handleDataSourceFileUpload(file, email, dsId)
} catch (error) {
  if (isDataSourceError(error)) {
    console.log('Error code:', error.code)
    console.log('User message:', error.userMessage)
  }
}
```

--------------------------------

### Google OAuth Client ID and Secret Creation

Source: https://github.com/xynehq/xyne/blob/main/docs/authentication/get-credentials.mdx

Instructions for creating OAuth client credentials within Google Cloud. This includes specifying the application type, authorized JavaScript origins, and redirect URIs for both local development and production environments.

```APIDOC
GoogleOAuthCredentials:
  CreateClient:
    - Type: Web Application
    - Name: [Your Client Name] (e.g., Xyne)
    - Authorized JavaScript Origins:
        - Local: http://localhost:3001
        - Production: http://<YOUR_AWS_EC2_IPv4_DNS>
        - Dev-Mode Note: http://localhost:3000
    - Authorized Redirect URIs:
        - Local: http://localhost:3001/v1/auth/callback
        - Production: https://<YOUR_AWS_EC2_IPv4_DNS>/v1/auth/callback
        - Local Alternative: http://localhost:3001/oauth/callback
        - Production Alternative: https://<YOUR_AWS_EC2_IPv4_DNS>/oauth/callback
        - Dev-Mode Note: http://localhost:3000/v1/auth/callback, http://localhost:3000/oauth/callback
  Output:
    - Client ID
    - Client Secret
  Usage:
    - Store in application's .env file.
```

--------------------------------

### Promtail Configuration for PM2 Logs

Source: https://github.com/xynehq/xyne/blob/main/observability/docs/observability.md

Configuration file for Promtail to scrape logs generated by PM2. It specifies the Loki endpoint and defines how to discover and label PM2 stdout and stderr log files.

```yaml
server:
  http_listen_port: 9080

positions:
  filename: /tmp/positions.yml

clients:
  - url: http://localhost:3100/loki/api/v1/push

scrape_configs:
- job_name: pm2-logs
  static_configs:
  - targets:
      - localhost
    labels:
      job: pm2
      type: stdout
    __path__: '~/.pm2/logs/*-out.log'

  - targets:
      - localhost
    labels:
      job: pm2
      type: stderr
    __path__: '~/.pm2/logs/*-error.log'
```

--------------------------------

### Google Workspace Delegation Scopes

Source: https://github.com/xynehq/xyne/blob/main/docs/authentication/service-accounts.mdx

A list of read-only scopes required for delegating access to Google Workspace services. These scopes are used when authorizing a service account's OAuth client ID in the Domain Wide Delegation settings.

```javascript
https://www.googleapis.com/auth/drive.readonly,
https://www.googleapis.com/auth/documents.readonly,
https://www.googleapis.com/auth/spreadsheets.readonly,
https://www.googleapis.com/auth/presentations.readonly,
https://www.googleapis.com/auth/contacts.readonly,
https://www.googleapis.com/auth/contacts.other.readonly,
https://www.googleapis.com/auth/gmail.readonly,
https://www.googleapis.com/auth/calendar.events.readonly,
https://www.googleapis.com/auth/admin.directory.user.readonly
```

--------------------------------

### Clone Xyne Repository

Source: https://github.com/xynehq/xyne/blob/main/docs/contribution/devmode.mdx

Clones the Xyne project repository from GitHub to your local machine. This is the first step in setting up the development environment.

```bash
git clone https://github.com/xynehq/xyne.git
```

--------------------------------

### Configure pm2-prom-module Port

Source: https://github.com/xynehq/xyne/blob/main/observability/docs/observability.md

Sets a custom port for the pm2-prom-module to expose metrics. The default port is 9988, but this command allows changing it to a specified value, like 10801.

```sh
pm2 set pm2-prom-module:port 10801
```

--------------------------------

### Configure Environment Variables

Source: https://github.com/xynehq/xyne/blob/main/docs/contribution/devmode.mdx

Sets up essential environment variables for the Xyne application, including API keys for various services (Google, AWS, OpenAI, Ollama, Together AI, Fireworks AI, Gemini) and database configurations. It's recommended to use a single `.env` file for clarity during development.

```env
ENCRYPTION_KEY=<YOUR_ENCRYPTION_KEY>
SERVICE_ACCOUNT_ENCRYPTION_KEY=<YOUR_SERVICE_ACCOUNT_ENCRYPTION_KEY>

GOOGLE_CLIENT_ID=<YOUR_GOOGLE_CLIENT_ID>
GOOGLE_CLIENT_SECRET=<YOUR_GOOGLE_CLIENT_SECRET>
GOOGLE_REDIRECT_URI=http://localhost:3000/v1/auth/callback

HOST=http://localhost:3000

JWT_SECRET=<YOUR_JWT_SECRET>

DATABASE_HOST=xyne-db
VESPA_HOST=vespa

## If using AWS Bedrock
AWS_ACCESS_KEY=<YOUR_AWS_ACCESS_KEY>
AWS_SECRET_KEY=<YOUR_AWS_ACCESS_SECRET>
AWS_REGION=<YOUR_AWS_REGION>
  
## OR [ If using Open AI ]
OPENAI_API_KEY=<YOUR_OPEN_API_KEY>

## OR [ If using Ollama ] 
OLLAMA_MODEL=<YOUR_OLLAMA_MODEL_NAME>

## OR [ If using Together AI ] 
TOGETHER_API_KEY=<YOUR_TOGETHER_API_KEY>
TOGETHER_MODEL=<YOUR_TOGETHER_MODEL>
TOGETHER_FAST_MODEL=<YOUR_TOGETHER_FAST_MODEL>

## OR [ If using Fireworks AI ] 
FIREWORKS_API_KEY=<YOUR_FIREWORKS_API_KEY>
FIREWORKS_MODEL=<YOUR_FIREWORKS_MODEL>
FIREWORKS_FAST_MODEL=<YOUR_FIREWORKS_FAST_MODEL>

## OR [If using Google AI]
GEMINI_API_KEY=<YOUR_GEMINI_API_KEY> 
GEMINI_MODEL=<YOUR_GEMINI_MODEL_NAME>

## If you are using custom OpenAI or Together AI endpoints
BASE_URL=<YOUR_BASE_URL>
```

--------------------------------

### Configure Production Environment Variables

Source: https://github.com/xynehq/xyne/blob/main/server/integrations/dataSource/README.md

Sets essential environment variables for the Xyne project's production environment. These variables control data source file size limits, conversion timeouts, temporary directory locations, and the path to the LibreOffice executable.

```bash
export DATASOURCE_MAX_FILE_SIZE_MB=50
export DATASOURCE_CONVERSION_TIMEOUT_MS=60000
export DATASOURCE_TEMP_DIR=/var/tmp/datasource
export LIBREOFFICE_PATH=/usr/bin/soffice
```

--------------------------------

### Restart Prometheus via Docker Compose

Source: https://github.com/xynehq/xyne/blob/main/observability/docs/observability.md

Restarts Prometheus instances managed by Docker Compose. This is necessary for Prometheus to pick up configuration changes, such as the new PM2 scraping job.

```sh
# If using docker-compose.metrics.yml
docker-compose -f deployment/docker-compose.metrics.yml restart prometheus

# Or if using docker-compose.selfhost.yml
docker-compose -f deployment/docker-compose.selfhost.yml restart prometheus
```

--------------------------------

### Prometheus Configuration for PM2 Metrics

Source: https://github.com/xynehq/xyne/blob/main/observability/docs/observability.md

Configures Prometheus to scrape metrics from the pm2-prom-module. This involves uncommenting a job definition in the Prometheus configuration file and specifying the target host and port.

```yaml
  - job_name: 'pm2'
    metrics_path: /
    scrape_interval: 2s
    static_configs:
      - targets: ['host.docker.internal:9988'] # Ensure this port matches the one pm2-prom-module is using (default is 9988)
```