# Run Python Workloads in a Remote Sandbox

Complex Python tasks often require libraries and an isolated runtime. This skill submits code to inference.sh and returns console output or generated files.

## Install

```bash
npx skillstore add 101-skills/python-executor
```

## Metadata

- Status: approved
- Slug: 101-skills-python-executor
- Skillstore revision: r1
- Version status: missing
- Tree hash: e7dfc159ab8ec1467df6a41d7d9453591d31715785fa4a066460349610403e6f
- Author: 101-skills
- GitHub username: 101-skills
- License: MIT
- Repository: https://github.com/101-skills/skills/tree/main/tools/utilities/python-executor/
- Ref: d71c7417a35d5c2624161bd2fe8de8a41a362128
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network
- Quality score: 50
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/101-skills-python-executor
- Manifest: https://skillstore.pages.dev/api/skills/101-skills-python-executor/manifest

## Capabilities

- Runs Python 3.10 code through the infsh/python-executor application.
- Processes and visualizes data with NumPy, Pandas, SciPy, Matplotlib, Seaborn, and Plotly.
- Makes HTTP requests and parses web content with common Python libraries.
- Creates image, video, audio, document, and 3D model files.
- Returns files written to the outputs directory with captured console output.

## Use Cases

- Analyze Operational Data: Clean a dataset, calculate summary metrics, and generate charts without preparing a local Python environment.
- Automate Web Research: Request approved public pages, parse selected fields, and return a structured summary.
- Generate Technical Assets: Create images, PDF reports, short videos, or 3D model exports from defined inputs.

## Prompt Templates

### Run a Simple Calculation

```
Use the Python executor to calculate [calculation]. Show the result and briefly explain the method.
```

### Summarize a Dataset

```
Analyze [dataset description] with Pandas. Report missing values, key statistics, and notable patterns. Do not include sensitive data.
```

### Create a Downloadable Report

```
Process [input data] and create a PDF report in the outputs directory. Include [sections] and charts for [metrics].
```

### Build a Controlled Research Pipeline

```
Request only [approved domains], enforce timeouts and response limits, extract [fields], validate results, and save the final table to outputs.
```

## Limitations

- Requires the belt CLI, an inference.sh account, and network access.
- Uses a CPU-only remote environment without GPU machine learning libraries.
- Supports non-interactive execution with a maximum documented timeout of 300 seconds.
- Submitted code and embedded data are processed by a third-party service.

## Best Practices

- Remove credentials and sensitive data before submitting code or inputs.
- Set the shortest practical timeout and write required artifacts to the outputs directory.
- Use approved domains, request timeouts, and response size limits for network tasks.

## Anti Patterns

- Do not place API keys, tokens, passwords, or private records inside submitted code.
- Do not scrape restricted sites or bypass access controls and usage policies.
- Do not rely on interactive windows, GPU acceleration, or unlimited execution time.

## Security Audit

- Audited at: 2026-07-12T10:13:22.264\+00:00
- Summary: The skill intentionally invokes the belt CLI, submits arbitrary Python to a third-party execution service, and demonstrates outbound HTTP requests. Most backtick detections are Markdown formatting, and the reconnaissance alert is unsupported, but remote execution, network access, and unpinned npx installation remain real risks.

## Stats

- Views: 0
- Downloads: 3
- Favorites: 0
- Popularity score: 0