viblr-sdk 1.1.0

Viblr SDK - Context as a Service for AI applications

Viblr SDK - Context as a Service

Perfect Prompts, First Try - Enhance your AI prompts with rich context from connected sources.

🚀 Quick Start

Installation

pip install -r requirements.txt

Basic Usage

from viblr_sdk import ViblrClient

# Initialize client
client = ViblrClient(api_key="your_api_key")

# Search for context
response = client.search(
    query="How to handle authentication errors?",
    max_results=5
)

print(f"Enhanced prompt: {response.enhanced_prompt}")
print(f"Citations: {response.search_results}")

🔧 Advanced Usage

Search with Filters

# Filter by specific sources
response = client.search(
    query="Recent changes to login system",
    sources=["gmail", "github"],
    content_types=["email", "code"],
    max_results=10
)

# Get raw results instead of enhanced prompt
raw_response = client.search(
    query="test query",
    output_format="raw_results"
)

Processing Options

# Enable advanced processing
response = client.search(
    query="Analyze customer feedback",
    processing={
        "sentiment": True,
        "keywords": True,
        "summary": True,
        "textrank": True  # For high-quality context (slower)
    }
)

print(f"Sentiment: {response.processing.sentiment}")
print(f"Keywords: {response.processing.keywords}")

Different Output Formats

# Enhanced prompt (default)
enhanced = client.search(query="test", output_format="enhanced_prompt")

# Raw search results
raw = client.search(query="test", output_format="raw_results")

# Processed results only
processed = client.search(query="test", output_format="processed_results")

🧪 Testing

Run the Example

python example_app.py

📚 API Reference

ViblrClient

search(query, max_results=5, sources=None, content_types=None, processing=None, output_format="enhanced_prompt")

Search for context and enhance prompts.

Parameters:

• query (str): The search query
• max_results (int): Maximum number of results to return (default: 5)
• sources (List[str], optional): Filter by sources (e.g., ['gmail', 'github'])
• content_types (List[str], optional): Filter by content types (e.g., ['email', 'code'])
• processing (Dict, optional): Processing options (sentiment, keywords, summary, textrank)
• output_format (str): Output format - 'enhanced_prompt', 'raw_results', 'processed_results'

Returns:

• ViblrResponse: Response object with enhanced_prompt, search_results, metadata

Raises:

• ViblrAPIError: API request failed
• ViblrAuthError: Authentication failed

ViblrResponse

Response object containing:

• enhanced_prompt (str): Enhanced prompt with context
• search_results (List[Dict]): Raw search results with citations
• status (str): Response status
• metadata (Dict): Additional metadata (intent analysis, processing results, etc.)

🎯 Use Cases

1. AI Chatbots

Enhance user prompts with project context before sending to AI models.

2. Code Review Assistants

Add code context, recent changes, and team discussions to review requests.

3. Customer Support

Enhance support tickets with customer history and product context.

4. Documentation Assistants

Add project documentation and code context to documentation requests.

🔒 Authentication

The SDK uses API key authentication. Get your API key from the Viblr dashboard.

📊 Context Sources

Supported context sources:

• Gmail: Email communications
• GitHub: Code repositories, issues, pull requests
• Slack: Team conversations and channels
• Jira: Project management and tickets
• Confluence: Documentation and knowledge base
• More coming soon...

🚀 Getting Started

1. Install the SDK: pip install -r requirements.txt
2. Get API Key: Sign up at Viblr dashboard
3. Connect Sources: Link your Gmail, GitHub, Slack, Jira, etc.
4. Search Context: Use the SDK in your AI applications

📞 Support

• Documentation: Viblr Docs
• Support: [e-mail]
• GitHub: Viblr SDK

🎉 Perfect Prompts, First Try

Stop explaining your work to AI. Let Viblr provide the context AI needs to give you amazing results immediately.