Agent Code Executor

Visualize AI writing and executing code in a sandboxed environment with real-time output, execution history, and security controls.

Features

Code Display: Syntax-highlighted editor view showing the agent's generated code
Output Panel: Live stdout, stderr, and return-value output from the sandboxed execution
Execution History: Browse previous runs with input snapshots and result diffs
Security Indicators: Visual badges showing sandbox isolation level and permission scope
Sandbox Controls: Start, stop, and restart execution with resource usage limits

Examples

Data Processing Script

Terminal — python3—python-sandbox-01

Sandboxed

python

10 lines

1import pandas as pd
2
3df = pd.read_csv("sales_q4.csv")
4summary = df.groupby("region").agg(
5    total_revenue=("revenue", "sum"),
6    avg_deal_size=("revenue", "mean"),
7    num_deals=("revenue", "count"),
8)
9summary["avg_deal_size"] = summary["avg_deal_size"].round(2)
10print(summary.to_markdown(tablefmt="grid"))

output

exit 0

+-----------+-----------------+-----------------+------------+

+===========+=================+=================+============+

| APAC | 384200 | 12806.7 | 30 |

| EMEA | 521800 | 14938.6 | 35 |

| NA | 697500 | 15500.0 | 45 |

+-----------+-----------------+-----------------+------------+

$ Process exited with code 0

Python

Sandboxed

1.24s

48.2 MB

Ln 10

Installation

npx shadcn add "https://agents-ui-kit.com/c/agent-code-executor.json"