# Run Python Code in a Remote Sandbox

Teams need quick Python execution without preparing local environments. This skill runs code through inference.sh with libraries for analysis, scraping, media, and automation.

## Install

```bash
npx skillstore add inference-sh/python-executor
```

## Metadata

- Status: approved
- Slug: inference-sh-python-executor
- Skillstore revision: r1
- Version status: missing
- Tree hash: 3047b05107ffca32587b2b54ae04eb4bd8bbd9856f8c4769d5acec88e3630260
- Author: inference-sh
- GitHub username: inference-sh
- License: MIT
- Repository: https://github.com/inference-sh/skills/tree/main/skills/python-executor/
- Ref: a06681402992ceae98ba04d54cfd4ab004862696
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: blocked
- Manual install advisory: allowed\_with\_warning
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network
- Quality score: 38
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/inference-sh-python-executor
- Manifest: https://skillstore.pages.dev/api/skills/inference-sh-python-executor/manifest

## Capabilities

- Executes Python 3.10 code through infsh/python-executor.
- Uses libraries such as pandas, matplotlib, Pillow, requests, Selenium, and Playwright.
- Returns files saved under outputs, including charts, CSVs, PDFs, videos, and 3D models.
- Supports timeout control from 1 to 300 seconds.
- Offers default 8GB RAM and high-memory 16GB RAM variants.
- Handles data analysis, scraping, media processing, PDF generation, and API calls.

## Use Cases

- Analyze Operational Data: Clean a dataset, compute summaries, and return charts or CSV outputs from a managed Python environment.
- Automate Public API Workflows: Call approved APIs, transform responses, and generate files without configuring a local Python runtime.
- Create Media And Documents: Generate images, simple videos, PDFs, and 3D files using preinstalled Python libraries.

## Prompt Templates

### Run a Small Python Task

```
Use python-executor to run a short Python task, explain the result, and return any saved output files.
```

### Analyze a Dataset

```
Use python-executor with pandas to clean the attached sales data, summarize trends, and save a chart under outputs.
```

### Process Media Files

```
Use python-executor to resize these images, build a contact sheet, and return the saved image files.
```

### Automate a Report Workflow

```
Use python-executor to fetch approved public API data, transform it, create a PDF report, and return generated files. Do not include secrets.
```

## Limitations

- Runs CPU-only and does not provide GPU or dedicated ML runtimes.
- Requires inference.sh CLI access and an authenticated account.
- Sends code and inputs to an external execution service.
- Non-interactive tasks should save files instead of opening UI windows.

## Best Practices

- Review code before execution and remove secrets, tokens, and private datasets.
- Set explicit timeouts and save generated files under outputs.
- Limit network calls to approved public URLs and follow site policies.

## Anti Patterns

- Piping remote installers directly into a shell in automated workflows.
- Running untrusted or private code without review.
- Using scraping examples against sites without permission.

## Security Audit

- Audited at: 2026-07-06T16:07:41.49\+00:00
- Summary: Most external command detections are false positives from Markdown code fences and inline package names. The audit confirms a critical pipe-to-shell installer command, intentional outbound network examples, and a high-risk remote arbitrary Python execution capability. No prompt injection attempt was found in SKILL.md.

## Stats

- Views: 286
- Downloads: 12
- Favorites: 0
- Popularity score: 0