# Process Images with Yescan OCR

Image documents often need OCR, cleanup, or conversion before use. This skill routes image files or URLs through the yescan CLI for OCR, enhancement, document export, and ID photo generation.

## Install

```bash
npx skillstore add yescan-ai/yescan-universal
```

## Metadata

- Status: approved
- Slug: yescan-ai-yescan-universal
- Version: 1.0.5
- Author version: 1.0.5
- Skillstore revision: r1
- Version status: valid
- Tree hash: f1a655ce61a4e38f1db61e09e7080cf58063438a3349f2a53dfd7b1ecd873e09
- Author: yescan-ai
- GitHub username: yescan-ai
- License: MIT
- Repository: https://github.com/yescan-ai/yescan-skill/tree/main/skills/yescan-universal/
- Ref: 64ca8af0f54a325752f08bd54e52151061ea659a
- 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: env\_access, network, filesystem, external\_commands
- Quality score: 38
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/yescan-ai-yescan-universal
- Manifest: https://skillstore.pages.dev/api/skills/yescan-ai-yescan-universal/manifest

## Capabilities

- Lists available yescan scenes and scene-specific parameters with yescan --list-scenes.
- Runs OCR for images, including IDs, invoices, tables, handwriting, and formulas.
- Enhances images with watermark, shadow, handwriting removal, correction, cropping, and HD enhancement.
- Converts image inputs to Word, Excel, or PDF through supported CLI scenes.
- Generates ID photo images and returns output file paths for generated assets.
- Handles local file paths or HTTP and HTTPS image URLs, with one input source per run.

## Use Cases

- Extract Text from Documents: Extract readable text from IDs, invoices, handwriting, formulas, and scanned pages.
- Clean Scanned Images: Remove shadows, marks, or watermarks from a user-provided image and return the saved result.
- Convert Image Tables: Turn table images or document photos into Word, Excel, or PDF outputs.

## Prompt Templates

### Run Basic OCR

```
Use yescan to extract readable text from this image: [path or URL]. Show me the recognized text.
```

### Improve a Document Image

```
Enhance this scanned image: [path]. Remove shadows and correct the page if supported. Return the output file path.
```

### Convert a Table Image

```
Convert this table image into an Excel document: [path]. Verify the conversion and give me the saved file path.
```

### Build a Multi-Step Workflow

```
Plan a yescan workflow for these images: [paths]. Identify needed scenes, ask about required parameters, run each step, and verify outputs.
```

## Limitations

- Requires the yescan Python package and a configured SCAN\_WEBSERVICE\_KEY.
- Sends user-selected images to Quark Scan services for processing.
- Handles images up to 5 MB and does not support video or pure text tasks.
- Does not accept base64 input directly; base64 images must be saved as files first.

## Best Practices

- Confirm the image source and task before sending data to the external service.
- Use current scene discovery output to choose scenes and parameters for the session.
- Review images for sensitive data before processing them with third-party OCR services.

## Anti Patterns

- Do not process images that the user did not explicitly provide.
- Do not guess scene names or parameter values without current CLI output.
- Do not report success until the CLI response and output path are verified.

## Security Audit

- Audited at: 2026-07-08T12:09:33.734\+00:00
- Summary: The reviewed files are readable documentation and skill instructions, not obfuscated payloads. The main confirmed risks are third-party image upload, API key handling, hidden credential storage, temp-file outputs, and yescan or pip command execution. I found no evidence of prompt injection, hidden endpoints, or attempts to bypass the audit.

## Stats

- Views: 0
- Downloads: 6
- Favorites: 0
- Popularity score: 0