An advanced AI-powered assistant for Odoo development that uses LangChain for modular LLM support, enabling easy switching between different language models.

• 🔄 Multi-LLM Support: Switch between Gemini, OpenAI, and Anthropic models
• 📋 Structured Prompts: LangChain PromptTemplates for consistent, maintainable prompts
• 🌐 Web Interface: Clean, responsive web UI with real-time updates
• 📱 CLI Interface: Command-line interface for direct terminal usage
• 🔧 Module Creation: Generate complete Odoo modules with best practices
• ⚙️ Module Customization: Modify existing modules intelligently
• 📊 Code Analysis: Comprehensive code quality and security analysis
• 🐛 Debug Assistant: AI-powered debugging with suggested fixes
• 🔒 Interactive Approval: Review all changes before applying
• 📦 File Backup: Automatic backup of modified files
• 🎯 Odoo-Specific: Tailored for Odoo 15/16 development patterns

• Python 3.8+
• UV package manager
• Odoo development environment
• API key for your preferred LLM provider

1. Clone and setup the project:

   cd /path/to/your/workspace
   git clone <your-repo> llama_agent
   cd llama_agent

2. Install dependencies with UV:

   uv sync

3. Configure environment: Copy .env.example to .env and configure:

   cp .env.example .env

4. Edit .env file:

   # LLM Configuration
   LLM_PROVIDER=gemini  # Options: gemini, openai, anthropic
   LLM_TEMPERATURE=0.1
   LLM_MAX_TOKENS=16384
   
   # Gemini API Configuration
   GEMINI_API_KEY=your_gemini_api_key_here
   
   # OpenAI API Configuration (optional)
   # OPENAI_API_KEY=your_openai_api_key_here
   # OPENAI_MODEL=gpt-4
   
   # Anthropic API Configuration (optional)
   # ANTHROPIC_API_KEY=your_anthropic_api_key_here
   # ANTHROPIC_MODEL=claude-3-5-sonnet-20241022
   
   # Odoo Development Settings
   ODOO_ADDONS_PATH=/path/to/your/odoo/addons
   ODOO_VERSION=16.0
   
   # Agent Configuration
   AGENT_MODE=development
   AUTO_APPROVE_CHANGES=false
   BACKUP_FILES=true

Start the web server:

uv run python web_ui.py

Then open http://localhost:8080 in your browser.

Web UI Features:

• 🔧 Create Module Tab: Generate new Odoo modules
• ⚙️ Customize Module Tab: Modify existing modules
• 📊 Analyze Code Tab: Get detailed code analysis
• 🐛 Debug Issue Tab: AI-powered debugging assistance

Run the command-line interface:

uv run python langchain_odoo_agent.py

CLI Options:

1. 🔍 Analyze existing code
2. 🔧 Create new module
3. ⚙️ Customize existing module
4. 🐛 Debug Odoo issue
5. ❌ Exit

LLM_PROVIDER=gemini
GEMINI_API_KEY=your_gemini_api_key

LLM_PROVIDER=openai
OPENAI_API_KEY=your_openai_api_key
OPENAI_MODEL=gpt-4  # or gpt-3.5-turbo

LLM_PROVIDER=anthropic
ANTHROPIC_API_KEY=your_anthropic_api_key
ANTHROPIC_MODEL=claude-3-5-sonnet-20241022

1. llm_config.py - LLM provider management and configuration
2. prompts.py - Structured prompt templates using LangChain
3. langchain_odoo_agent.py - Main agent logic with LangChain integration
4. web_ui.py - Flask-based web interface
5. templates/index.html - Responsive web UI with Alpine.js

from langchain_core.prompts import PromptTemplate
from langchain_google_genai import ChatGoogleGenerativeAI
from llm_config import LLMConfig
from prompts import get_prompt_template

# Initialize LLM
llm_config = LLMConfig(provider="gemini")
llm = llm_config.get_llm()

# Get prompt template
prompt_template = get_prompt_template("module_creation")

# Create chain
chain = prompt_template | llm | str_parser

# Execute
response = chain.invoke({
    "odoo_version": "16.0",
    "module_name": "my_module",
    "description": "My custom module",
    "features": "Custom functionality"
})

All prompts are now managed as LangChain PromptTemplate objects:

• CODE_ANALYSIS_TEMPLATE - For code quality analysis
• MODULE_CREATION_TEMPLATE - For generating new modules
• MODULE_CUSTOMIZATION_TEMPLATE - For modifying existing modules
• DEBUG_ISSUE_TEMPLATE - For debugging assistance

llama_agent/
├── llm_config.py           # LLM provider configuration
├── prompts.py              # LangChain prompt templates
├── langchain_odoo_agent.py # Main agent with LangChain
├── web_ui.py               # Flask web interface
├── templates/
│   └── index.html          # Web UI template
├── pyproject.toml          # UV project configuration
├── .env                    # Environment configuration
├── .env.example            # Environment template
└── README.md               # This file

1. Update LLMProvider enum in llm_config.py:

   class LLMProvider(Enum):
       GEMINI = "gemini"
       OPENAI = "openai"
       ANTHROPIC = "anthropic"
       NEW_PROVIDER = "new_provider"  # Add here

2. Add provider method in LLMConfig class:

   def _get_new_provider_llm(self) -> BaseLLM:
       api_key = os.getenv("NEW_PROVIDER_API_KEY")
       if not api_key:
           raise ValueError("NEW_PROVIDER_API_KEY not found")
       
       return NewProviderLLM(
           api_key=api_key,
           temperature=self.temperature,
           max_tokens=self.max_tokens
       )

3. Update get_llm() method to handle the new provider.

1. Define the template in prompts.py:

   NEW_TEMPLATE = PromptTemplate(
       input_variables=["param1", "param2"],
       template="""Your prompt template here with {param1} and {param2}"""
   )

2. Register in get_prompt_template() function.

3. Use in agent methods:

   prompt_template = get_prompt_template("new_template")
   chain = prompt_template | self.llm | self.str_parser
   response = chain.invoke({"param1": value1, "param2": value2})

# CLI
uv run python langchain_odoo_agent.py

# Web UI
uv run python web_ui.py
# Then visit http://localhost:8080

1. Test with Gemini:

   export LLM_PROVIDER=gemini
   uv run python langchain_odoo_agent.py

2. Test with OpenAI:

   export LLM_PROVIDER=openai
   export OPENAI_API_KEY=your_key
   uv run python langchain_odoo_agent.py

3. Test with Anthropic:

   export LLM_PROVIDER=anthropic
   export ANTHROPIC_API_KEY=your_key
   uv run python langchain_odoo_agent.py

1. "LLM provider not found"

   • Check LLM_PROVIDER value in .env
   • Ensure the provider's API key is configured

2. "Failed to parse JSON response"

   • LLM response may be truncated
   • Try increasing LLM_MAX_TOKENS
   • Check LLM provider rate limits

3. "Module path not found"

   • Verify ODOO_ADDONS_PATH points to correct directory
   • Ensure path permissions are correct

4. Web UI not loading

   • Check if port 8080 is available
   • Try changing UI_PORT in .env
   • Check firewall settings

Enable detailed logging:

export UI_DEBUG=true
uv run python web_ui.py

This project is licensed under the MIT License.

1. Fork the repository
2. Create a feature branch
3. Add your changes with tests
4. Submit a pull request

For issues and questions:

• Open an issue on GitHub
• Check the troubleshooting section
• Review the LangChain documentation for advanced usage

Happy Odoo Development! 🎉