run-python

Execute one-shot Python 3 scripts in an isolated sandbox with audit archiving.

Updated May 7, 2026

Search Tags:#automation #security #python #audit #sandbox #execute #ad-hoc

Ready to run

npx skills add https://github.com/gongzeq/VAPT3 --skill run-python

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: run-python
Source: https://github.com/gongzeq/VAPT3/tree/main/secbot/skills/run-python
Command: npx skills add https://github.com/gongzeq/VAPT3 --skill run-python

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Execute short, one-shot Python 3 scripts within the secbot sandbox to perform quick transforms, probes, PoCs, and debugging without leaving the workspace.

Core Features & Use Cases

Isolated interpreter using python3 with -I -B to prevent Pycache and env leakage.
Archives the script under <scan_dir>/run-python with a timestamp for auditability.
Enforces max code size and timeout, returns stdout_tail and exit status for deterministic results.
Supports quick ad-hoc automation such as data shaping, protocol probes, and PoC testing within secured workflows.

Quick Start

Provide a small Python script in the code field to execute once inside the sandbox with a 60-second default timeout.

Frequently Asked Questions about run-python

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I run ad-hoc Python scripts safely inside a sandbox for quick debugging?▼

To run ad-hoc Python safely, execute short one-shot scripts inside an isolated interpreter using -I -B flags to prevent environment leakage and Pycache generation, returning stdout tail and exit status for deterministic results.

What is the best way to execute quick data transforms without leaving my secured workflow?▼

The best way to execute quick data transforms is running a small Python 3 script in the secbot sandbox, which caps code size and timeout to ensure lightweight automation within secured workflows.

Do I need approval before running the first Python script in the sandbox?▼

Yes, you need a HighRiskGate approval before first use to execute Python inside the sandbox, ensuring ad-hoc tasks meet security requirements before running.

Can I use this for PoC debugging and protocol probing with a time limit?▼

Yes, you can use it for PoC debugging and protocol probing with a 60-second default timeout, which caps execution time to ensure deterministic results for quick probes.

Are there limitations when running Python scripts with isolated interpreter flags?▼

Limitations include a capped code size and a 60-second default timeout, and the isolated interpreter uses -I -B flags to prevent Pycache and environment leakage, restricting long or heavy scripts.