Vanaras Agent Framework

🌟 Features

🚀 Quick Start

Installation

pip install vanaras-agent-framework

Create Your First Agent

from vanaras.core.agent import Agent
from vanaras.core.task import Task
from vanaras.core.crew import Crew
# Define an agent
coder = Agent(
 role="Python Developer",
 goal="Write clean, efficient Python code",
 backstory="Expert Python developer with 10 years of experience"
)
# Create a task
task = Task(
 description="Create a function to calculate fibonacci numbers",
 expected_output="A working Python function",
 agent=coder
)
# Execute
crew = Crew(agents=[coder], tasks=[task])
result = crew.kickoff()
print(result)

Use Templates

# Initialize from a template
vanaras init code_analyzer
# List available templates
vanaras init --list
# Describe a template
vanaras init --describe code_analyzer

🏗️ Ecosystem

1. Core Framework (this repo)

2. Vanaras Flow

3. VS Code Extension

📖 Documentation

Agent System

from vanaras.core.agent import Agent
agent = Agent(
 role="Data Analyst",
 goal="Analyze data and provide insights",
 backstory="Expert in data analysis and visualization",
 tools=["read_file", "write_file"] # Optional custom tools
)

Crew Orchestration

from vanaras.core.crew import Crew, Process
# Sequential execution
crew = Crew(
 agents=[agent1, agent2],
 tasks=[task1, task2],
 process=Process.SEQUENTIAL
)
# DAG execution (parallel where possible)
crew = Crew(
 agents=[agent1, agent2, agent3],
 tasks=[task1, task2, task3],
 process=Process.DAG
)
# Hierarchical (manager coordinates workers)
crew = Crew(
 agents=[worker1, worker2],
 tasks=[task1, task2],
 process=Process.HIERARCHICAL,
 manager_llm="gpt-4"
)

Python Bridge (API Server)

# Start the bridge
vanaras serve --port 5001
# Test it
curl http://localhost:5001/health
curl -X POST http://localhost:5001/chat \
 -H "Content-Type: application/json" \
 -d '{"message": "Create a hello world function"}'

Built-in Tools

Custom Tools

from vanaras.tools.registry import registry, ToolMetadata, ToolCategory
def my_custom_tool(arg1, arg2):
 """Your custom tool logic."""
 return f"Processed {arg1} and {arg2}"
registry.register(ToolMetadata(
 name="my_tool",
 description="Description of what it does",
 category=ToolCategory.CUSTOM,
 function=my_custom_tool,
 destructive=False
))

🎯 Use Cases

🔧 CLI Commands

# Initialize from template
vanaras init <template_name>
# Run a task
vanaras run <task_file>
# Start Python Bridge
vanaras serve [--port PORT] [--host HOST]
# Launch UI Dashboard
vanaras dashboard [--port PORT]
# Tool registry
vanaras tools list
vanaras tools describe <tool_name>

🏃 Examples

Multi-Agent Workflow

from vanaras.core.agent import Agent
from vanaras.core.task import Task
from vanaras.core.crew import Crew, Process
# Define agents
researcher = Agent(
 role="Researcher",
 goal="Research and gather information",
 backstory="Expert researcher with attention to detail"
)
writer = Agent(
 role="Writer",
 goal="Write clear, engaging content",
 backstory="Professional technical writer"
)
# Define tasks
research_task = Task(
 description="Research the topic of AI agents",
 expected_output="A summary of key findings",
 agent=researcher
)
writing_task = Task(
 description="Write a blog post based on the research",
 expected_output="A well-written blog post",
 agent=writer,
 context=[research_task] # Depends on research_task
)
# Execute
crew = Crew(
 agents=[researcher, writer],
 tasks=[research_task, writing_task],
 process=Process.SEQUENTIAL,
 verbose=True
)
result = crew.kickoff()

🛠️ Development

Setup

git clone https://github.com/vanarasai/agent-framework.git
cd agent-framework
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
pip install -e .

Run Tests

pytest tests/

Build Package

python -m build

📊 Metrics & Monitoring

🔒 Security

🤝 Contributing

📄 License

🔗 Links

💬 Support