891,873 AI resources indexedWorking toward the world’s largest AI resource index.

PowerPoint Shell Verification

Verify PowerPoint presentation contents using python-pptx via shell when standard file readers fail

Installable skillMITBy AG064Updated Oct 30, 2026
INTENDED FOROpenClawInstall in a listed agent.

instructions.md

---
name: pptx-shell-verify-38bda23a5c4c
description: Verify PowerPoint presentation contents using python-pptx via shell when standard file readers fail
---

# PowerPoint Shell Verification

## When to Use This Skill

Use this skill when:
- `read_file` or other built-in tools fail to read PowerPoint (.pptx) files
- You need to verify presentation structure (slide count, titles, content)
- Standard file inspection methods return errors or incomplete data

## Prerequisites

- Python 3.x installed in the execution environment
- `python-pptx` library available (install if needed)

## Step-by-Step Instructions

### Step 1: Attempt Standard File Read

First, try to read the file using standard methods:

```
read_file: path/to/presentation.pptx
```

If this fails or returns unusable data, proceed to Step 2.

### Step 2: Verify python-pptx Availability

Check if the library is installed:

```
run_shell: python -c "import pptx; print('pptx available')"
```

If not available, install it:

```
run_shell: pip install python-pptx
```

### Step 3: Create Verification Script

Create a Python script to extract presentation metadata:

```python
# verify_pptx.py
from pptx import Presentation
import sys
import json

def verify_presentation(filepath):
    try:
        prs = Presentation(filepath)
        data = {
            "slide_count": len(prs.slides),
            "slides": []
        }
        
        for i, slide in enumerate(prs.slides):
            slide_data = {
                "slide_number": i + 1,
                "title": None,
                "shapes_count": len(slide.shapes)
            }
            
            # Extract title if present
            if slide.shapes.title:
                slide_data["title"] = slide.shapes.title.text
            
            data["slides"].append(slide_data)
        
        print(json.dumps(data, indent=2))
        return True
    except Exception as e:
        print(f"ERROR: {str(e)}", file=sys.stderr)
        return False

if __name__ == "__main__":
    if len(sys.argv) < 2:
        print("Usage: python verify_pptx.py <file.pptx>", file=sys.stderr)
        sys.exit(1)
    
    success = verify_presentation(sys.argv[1])
    sys.exit(0 if success else 1)
```

### Step 4: Execute Verification

Run the verification script:

```
run_shell: python verify_pptx.py path/to/presentation.pptx
```

### Step 5: Interpret Results

Expected output format:

```json
{
  "slide_count": 10,
  "slides": [
    {
      "slide_number": 1,
      "title": "Introduction",
      "shapes_count": 3
    },
    ...
  ]
}
```

Use this data to:
- Confirm expected slide count matches
- Verify slide titles are present
- Validate presentation structure before delivery

## Example Validation Checks

```python
# Check slide count matches requirement
assert data["slide_count"] >= expected_min_slides

# Check all slides have titles
for slide in data["slides"]:
    assert slide["title"] is not None, f"Slide {slide['slide_number']} missing title"
```

## Troubleshooting

| Issue | Solution |
|-------|----------|
| Corrupted file | Verify file was created successfully, check file size |
| Permission denied | Ensure file path is accessible, check read permissions |
| Library not found | Install with `pip install python-pptx` |
| Empty presentation | File may be 0 bytes or incomplete write |

## Best Practices

1. **Always verify after creation**: Run verification immediately after generating a PowerPoint file
2. **Keep verification script reusable**: Store the script for future presentations
3. **Validate before delivery**: Never deliver a presentation without confirming structure
4. **Log verification results**: Keep verification output for audit trails

Platform guidance

Imported from the listed AgentSkill.sh category. Complete supplied package files and notices are retained. Required external tools and credentials are not installed. No bundled code was executed during import. Individual agent task behavior has not been tested.