testteller 2.0.2

AI-Powered Test Agent for Test Case Generation and Automation using RAG & GenAI

TestTeller Agent

TestTeller is the AI-powered Test agent that transforms your documentation into comprehensive test suites and executable automation code. Powered by a dual-feedback RAG architecture with support for multiple GenAI/LLMs (Google Gemini, OpenAI, Anthropic Claude, Local Llama), TestTeller analyzes your requirements, designs, and existing code to generate strategic test cases and automate them across multiple programming languages and supported testing frameworks.

Why TestTeller?

TestTeller transforms documentation and code into comprehensive test strategies and executable automation. Unlike traditional testing tools, TestTeller uses dual-feedback RAG architecture to understand your requirements and generate intelligent test scenarios.

Test Types Generated

• End-to-End (E2E) Tests: Complete user journeys across frontend, middleware, and backend services
• Integration Tests: Component integration (FE-BE, service-to-service, event-driven) and contract validation
• Technical Tests: Performance, security, resilience testing with infrastructure focus
• Mocked System Tests: Isolated component testing with mocked dependencies

Critical Features

• Dual-Feedback RAG Enhancement: Self-improving system that learns from generation cycles and stores high-quality outputs
• Multi-Provider LLM Support: Works with Google Gemini, OpenAI, Anthropic Claude, and local Llama/Ollama
• Universal Document Intelligence: Advanced parsing for PDFs, DOCX, XLSX, MD, TXT with context understanding
• Code Repository Analysis: Ingests and analyzes code from GitHub repos or local folders.

Supported Languages & Frameworks

Test Generation: All test types with tabular summaries and detailed specifications
Automation: Python (pytest, unittest), JavaScript/TypeScript (Jest, Mocha, Cypress, Playwright), Java (JUnit, TestNG), and more

Real-world workflow: Ingest project (PRDs/Contracts/Design/Schema etc.) documentation & project code → Generate strategic test cases covering authentication, error handling, and edge cases → Create executable Selenium/Playwright automation with proper setup and assertions → Commit your code.

Key Features

• 🤖 Generator Agent: Virtual Test architect with dual-feedback RAG enhancement - analyzes docs and generates strategic test cases with intelligent categorization (E2E, integration, security, edge cases)
• ⚡ Automator Agent: Multi-language code generation across Python, JavaScript, TypeScript, Java with 20+ framework support (pytest, Jest, JUnit, Playwright, Cypress, Cucumber, etc.)
• 🔧 Multi-Provider GenAI/LLM: Choose your AI provider - Google Gemini, OpenAI, Anthropic Claude, or run completely local with Llama/Ollama
• 📄 Universal Document Intelligence: Advanced RAG ingestion for PDFs, DOCX, XLSX, MD, TXT - understands context and generates appropriate test focus
• 🔄 Self-Learning System: Dual-feedback architecture gets smarter with each use - stores high-quality outputs and learns from automation success patterns

→ View Detailed Features | Technical Architecture

Quick Start

Prerequisites

• Python 3.11+
• API key for at least one LLM provider:
  • Google Gemini (recommended)
  • OpenAI
  • Anthropic Claude
  • Ollama (local or within accessible environment)

Installation

# Install from PyPI
pip install testteller

# Or install from source
git clone https://github.com/iAviPro/testteller-agent.git
cd testteller-agent
pip install -e .

Basic Usage - Get Started in 2 Minutes

# 1. Configure your LLM provider (interactive wizard)
testteller configure

# 2. Ingest your documentation (supports PDF, DOCX, XLSX, MD, TXT)
testteller ingest-docs requirements.pdf --collection-name my_project

# 3. Ingest code from repository or local folder
testteller ingest-code https://github.com/user/repo --collection-name my_project
# OR: testteller ingest-code ./src --collection-name my_project

# 4. Generate strategic test cases with RAG context
testteller generate "Create comprehensive API integration tests" --collection-name my_project --output-file tests.pdf

# 5. Generate executable automation code
testteller automate tests.pdf --language python --framework pytest --output-dir ./tests

Enhanced Examples:

# E2E Testing Workflow
testteller ingest-docs user_stories.pdf --collection-name webapp
testteller ingest-code ./frontend --collection-name webapp  
testteller generate "E2E user registration and checkout flow" --collection-name webapp
testteller automate output.pdf --language javascript --framework cypress

# API Testing with Security Focus
testteller ingest-docs api_spec.pdf --collection-name api
testteller generate "API security and integration tests" --collection-name api --output-format pdf
testteller automate tests.pdf --language python --framework pytest

# Microservices Testing
testteller ingest-code ./services --collection-name microservices
testteller generate "Inter-service communication and resilience tests" --collection-name microservices
testteller automate output.pdf --language java --framework junit

What happens? TestTeller's dual-feedback RAG analyzes your ingested docs and code, generates strategic test cases using structured templates (E2E, Integration, Technical, Mocked), then creates production-ready automation with proper setup, data management, and CI/CD integration.

Try TestTeller Now

No API Keys? No problem - use local Llama:

# Install Ollama (macOS/Linux)  
curl -fsSL https://ollama.ai/install.sh | sh
ollama pull llama3.2

# Configure TestTeller for local use
testteller configure --provider llama

Docker Support

# Clone and setup
git clone https://github.com/iAviPro/testteller-agent.git
cd testteller-agent
cp .env.example .env  # Add your API keys
docker-compose up -d

# Use with Docker
docker-compose exec app testteller configure
docker-compose exec app testteller ingest-docs document.pdf --collection-name project

Documentation

• Complete Features - Detailed feature descriptions and capabilities
• Technical Architecture - System design and technical details
• Command Reference - Complete CLI command documentation
• Testing Guide - Test suite and validation documentation

Common Issues

Run testteller configure if you encounter API key errors. For Docker issues, check logs with docker-compose logs app.

---

License

This project is licensed under the Apache-2.0 License - see the LICENSE file for details.