# Run Python Tasks in a Remote Sandbox

Many AI workflows need short Python jobs without local setup. This skill runs Python through inference.sh with common analysis, scraping, media, and document libraries.

## Install

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

## Metadata

- Status: approved
- Slug: inference-sh-skills-python-executor
- Skillstore revision: r1
- Version status: missing
- Tree hash: e7dfc159ab8ec1467df6a41d7d9453591d31715785fa4a066460349610403e6f
- Author: inference-sh-skills
- GitHub username: inference-sh-skills
- License: MIT
- Repository: https://github.com/inference-sh-skills/skills/tree/main/tools/utilities/python-executor/
- Ref: a25199bc7d6b82598536822d1738eb5d5f54025b
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: confirmation\_required
- Manual install advisory: allowed
- 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-skills-python-executor
- Manifest: https://skillstore.pages.dev/api/skills/inference-sh-skills-python-executor/manifest

## Capabilities

- Runs Python snippets with configurable timeout through belt.
- Uses preinstalled libraries for data, scraping, browser automation, images, video, documents, and 3D files.
- Returns files saved under outputs, including charts, CSVs, images, videos, PDFs, and 3D models.
- Supports data analysis, visualization, web scraping, API calls, media processing, and document generation.
- Offers default and high-memory variants for larger workloads.

## Use Cases

- Analyze data tables: Process CSV or pasted data, calculate summaries, and return charts or transformed files.
- Automate approved web tasks: Fetch approved pages or APIs, extract structured information, and save results as files.
- Generate media and documents: Create images, videos, PDFs, and 3D assets from short Python scripts.

## Prompt Templates

### Run a small calculation

```
Use python-executor to run a short Python script that summarizes this data: [paste data]. Return the printed result and any output files.
```

### Create a chart

```
Use python-executor to build a chart from this table: [table]. Save the chart to outputs/chart.png and summarize the result.
```

### Scrape an approved page

```
Use python-executor to fetch [approved URL], extract [fields], and save a CSV in outputs. Use a timeout of [seconds].
```

### Build a processing pipeline

```
Use python-executor to process [input files], call [approved API], generate [report type], and return all files saved in outputs.
```

## Limitations

- Requires the inference.sh belt CLI and login before use.
- Runs CPU-only jobs and does not provide GPU acceleration.
- Code execution is non-interactive, so scripts must save outputs instead of opening windows.
- User code may access networks if written to do so; review code before execution.

## Best Practices

- Review generated Python before execution, especially network requests and file output paths.
- Keep secrets out of prompts and code sent to remote execution.
- Set practical timeouts and save intentional artifacts under outputs.

## Anti Patterns

- Running unreviewed code from unknown sources.
- Sending private data, API keys, or credentials to the executor.
- Using the skill for long-running, interactive, GPU, or production workloads.

## Security Audit

- Audited at: 2026-07-05T17:54:20.543\+00:00
- Summary: Most static shell findings are Markdown backticks or examples, but the skill intentionally routes code through the external belt CLI and inference.sh. Confirmed risks include remote arbitrary Python execution, outbound network examples, mutable install dependencies, and automatic returned output files; no prompt injection text was found.

## Stats

- Views: 59
- Downloads: 8
- Favorites: 0
- Popularity score: 0