aiwebexplorer 0.0.4

An agent for agents to explore the web

AIWebExplorer 🌐

An agent for agents to explore the web

📦 Installation

This project uses uv for dependency management.

# Clone the repository
git clone <repository-url>
cd AIWebExplorer

# Install dependencies
uv sync

# Activate virtual environment
source .venv/bin/activate

🛠️ Development

# Run linting
uv run ruff check .

# Run formatting
uv run ruff format .

# Run type checking
uv run ruff check --select I

⚙️ Environment Variables

Copy .env.example to .env and adjust the values:

# Environment setting (DEV, TEST, CI, PROD)
AWE_ENV=DEV

# Logging level (DEBUG, INFO, WARNING, ERROR, CRITICAL)
AWE_LOG_LEVEL=INFO

# LLM Provider configuration (REQUIRED)
# Supported providers: openai, togetherai, deepseek
AWE_LLM_PROVIDER=openai

# LLM Model to use (REQUIRED)
# Examples:
# - OpenAI: gpt-4, gpt-4-turbo, gpt-3.5-turbo
# - TogetherAI: meta-llama/Llama-2-70b-chat-hf, mistralai/Mixtral-8x7B-Instruct-v0.1
# - DeepSeek: deepseek-chat, deepseek-coder
AWE_LLM_MODEL=gpt-4

# API Key for the selected provider (REQUIRED)
AWE_LLM_API_KEY=your-api-key-here

Configuration Options

• AWE_ENV: Application environment (default: DEV)
  • Options: DEV, TEST, CI, PROD
• AWE_LOG_LEVEL: Logging verbosity level (default: INFO)
  • Options: DEBUG, INFO, WARNING, ERROR, CRITICAL
• AWE_LLM_PROVIDER: LLM provider selection (REQUIRED)
  • Supported providers: openai, togetherai, deepseek
  • Can be overridden when creating agents programmatically
• AWE_LLM_MODEL: Model identifier to use (REQUIRED)
  • Must be compatible with the selected provider
  • Can be overridden when creating agents programmatically
• AWE_LLM_API_KEY: API key for authentication (REQUIRED)
  • Must be valid for the selected provider
  • Can be overridden when creating agents programmatically

Supported Providers

🤖 OpenAI

• Provider: openai
• Models: gpt-4, gpt-4-turbo, gpt-3.5-turbo, and more
• API Key: Get from OpenAI Platform

🔗 TogetherAI

• Provider: togetherai
• Models: Various open-source models including Llama, Mixtral, etc.
• API Key: Get from Together.ai

🔍 DeepSeek

• Provider: deepseek
• Models: deepseek-chat, deepseek-coder
• API Key: Get from DeepSeek Platform

🧪 Testing

For comprehensive testing documentation, including how to run tests, use dependency injection for mocking, and write new tests, see the Tests README.

Running Tests

# Run all tests
pytest

# Run with verbose output
pytest -v

# Run specific test file
pytest tests/test_webexplorer_integration.py

📊 Evaluation Reports

Performance evaluation reports are available in the tests/reports/ directory:

• Amazon Extraction Report - Evaluation of product information extraction from Amazon
• Wikipedia Extraction Report - Evaluation of information extraction from Wikipedia

These reports track the accuracy and performance of the WebExplorer across different types of websites and extraction tasks.

✨ New Features

To develop a new feature:

1. Create a feature branch from develop:

   git checkout develop
   git pull origin develop
   git checkout -b feature/your-feature-name
   

2. Work on your feature and commit changes:

   git add .
   git commit -m "feat: add your new feature"
   git push origin feature/your-feature-name
   

3. Create a Pull Request to develop branch

4. After review and merge, delete the feature branch

🚀 New Versions

Option 1: Automated Release (Recommended)

For automated releases, simply commit with the release message:

git commit -m "chore: release v1.2.0"
git push origin master

This will automatically:

• Create the version tag
• Publish to PyPI
• Create a GitHub release

Option 2: Manual Release

1. Create a release branch from develop:

   git checkout develop
   git pull origin develop
   git checkout -b release/v1.2.0
   

2. Update CHANGELOG.md with your changes

3. Merge to master and create version tag:

   git checkout master
   git merge release/v1.2.0
   git tag v1.2.0
   git push origin master --tags
   

4. Merge back to develop:

   git checkout develop
   git merge release/v1.2.0
   git push origin develop
   

The CI/CD pipeline will automatically:

• Run tests and linting
• Build and publish to PyPI when version tags are pushed
• Create GitHub releases

Version numbering:

• Patch (1.0.0 → 1.0.1): Bug fixes
• Minor (1.0.0 → 1.1.0): New features
• Major (1.0.0 → 2.0.0): Breaking changes

License

[Add your license here]