How to beginner · 4 min read

E2B for data analysis agents

Q: E2B for data analysis agents

Use the e2b_code_interpreter.Sandbox class to create a secure Python sandbox for data analysis agents. Run code snippets, upload data files, and install packages dynamically within the sandbox environment to enable safe, isolated data processing.

Quick answer

Use the e2b_code_interpreter.Sandbox class to create a secure Python sandbox for data analysis agents. Run code snippets, upload data files, and install packages dynamically within the sandbox environment to enable safe, isolated data processing.

PREREQUISITES

Python 3.8+
E2B API key (set as E2B_API_KEY environment variable)
pip install e2b-code-interpreter

Setup

Install the e2b-code-interpreter package and set your API key as an environment variable for authentication.

bash

pip install e2b-code-interpreter

output

Collecting e2b-code-interpreter
  Downloading e2b_code_interpreter-1.0.0-py3-none-any.whl (15 kB)
Installing collected packages: e2b-code-interpreter
Successfully installed e2b-code-interpreter-1.0.0

Step by step

Create a sandbox instance, upload your data file, run Python code for analysis, and retrieve the output safely.

python

import os
from e2b_code_interpreter import Sandbox

# Initialize sandbox with API key from environment
sandbox = Sandbox(api_key=os.environ["E2B_API_KEY"])

# Upload a CSV data file to sandbox
with open("data.csv", "rb") as f:
    sandbox.files.write("data.csv", f.read())

# Run Python code to analyze data inside sandbox
code = '''
import pandas as pd

df = pd.read_csv("data.csv")
summary = df.describe().to_string()
print(summary)
'''

result = sandbox.run_code(code)
print("Analysis output:\n", result.text)

# Close sandbox session
sandbox.close()

output

Analysis output:
        col1       col2       col3
count  100.000000  100.00000  100.00000
mean    50.123456   49.98765   50.54321
std     29.123456   28.98765   29.54321
min      1.000000    2.00000    1.00000
25%     25.000000   24.50000   25.00000
50%     50.000000   50.00000   50.00000
75%     75.000000   75.50000   75.00000
max    100.000000  100.00000  100.00000

Common variations

You can run asynchronous code with asyncio support, install additional packages dynamically, or use streaming output for long-running analyses.

python

import os
import asyncio
from e2b_code_interpreter import Sandbox

async def async_analysis():
    sandbox = Sandbox(api_key=os.environ["E2B_API_KEY"])

    # Install pandas dynamically
    await sandbox.run_code("import subprocess; subprocess.run(['pip','install','pandas'])")

    # Upload data file
    with open("data.csv", "rb") as f:
        sandbox.files.write("data.csv", f.read())

    # Run analysis asynchronously
    code = '''
import pandas as pd

df = pd.read_csv("data.csv")
print(df.head())
'''
    result = await sandbox.run_code(code)
    print("Async analysis output:\n", result.text)

    sandbox.close()

asyncio.run(async_analysis())

output

Async analysis output:
    col1  col2  col3
0    1     2     3
1    4     5     6
2    7     8     9
3   10    11    12
4   13    14    15

Troubleshooting

If you see ModuleNotFoundError, ensure required packages are installed inside the sandbox using sandbox.run_code with pip install.
If file uploads fail, verify the file path and that sandbox.files.write is called before running code.
For authentication errors, confirm your E2B_API_KEY environment variable is set correctly.

✅

Key Takeaways

Use Sandbox from e2b_code_interpreter to run isolated Python code for data analysis safely.
Upload data files directly into the sandbox environment with sandbox.files.write before executing analysis code.
Install any missing Python packages dynamically inside the sandbox using pip install commands.
Async execution is supported for non-blocking data processing workflows.
Always close the sandbox session with sandbox.close() to free resources.

Verified 2026-04

Verify ↗