> ## Documentation Index
> Fetch the complete documentation index at: https://praison.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.
# Python
> Execute Python code
## Overview
Python tool allows you to execute Python code from your AI agents.
The user sends code; the agent runs it in a sandbox and returns stdout and results.
```mermaid theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
graph LR
subgraph "PythonTool Flow"
User[👤 User] --> Agent[🤖 Agent]
Agent --> Tool[🔧 PythonTool]
Tool --> Result[✅ Result]
Result --> Agent
end
classDef agent fill:#8B0000,stroke:#7C90A0,color:#fff
classDef tool fill:#189AB4,stroke:#7C90A0,color:#fff
classDef output fill:#10B981,stroke:#7C90A0,color:#fff
class User,Agent agent
class Tool tool
class Result output
```
## Installation
```bash theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
pip install "praisonai[tools]"
```
## Quick Start
```python theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
from praisonai_tools import PythonTool
# Initialize
python = PythonTool()
# Execute code
result = python.execute("print(2 + 2)")
print(result)
```
Use the same tool with an agent — see **Usage with Agent** below, or pass env vars and options from the sections above.
## Usage with Agent
```python theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
from praisonaiagents import Agent
from praisonai_tools import PythonTool
agent = Agent(
name="CodeRunner",
instructions="You help execute Python code.",
tools=[PythonTool()]
)
response = agent.chat("Calculate the factorial of 10")
print(response)
```
## Available Methods
### execute(code)
Execute Python code.
```python theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
from praisonai_tools import PythonTool
python = PythonTool()
result = python.execute('''
import math
print(math.factorial(10))
''')
```
## Security Warning
⚠️ **Use with caution!** Executing arbitrary code can be dangerous.
## How It Works
```mermaid theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
sequenceDiagram
participant User
participant Agent
participant Tool as Tool
participant Svc as Python runtime
User->>Agent: Request
Agent->>Tool: Call tool
Tool->>Svc: Query
Svc-->>Tool: Data
Tool-->>Agent: Result
Agent-->>User: Response
```
***
## Best Practices
Only execute code generated from trusted instructions. Treat all output as untrusted.
Wrap execution so tracebacks are returned to the agent instead of terminating the run.
Pass all inputs explicitly — don't rely on state persisting between executions.
***
## Related Tools
Shell commands
Math calculations