Skills video-enhancement
🎬

video-enhancement

Medium Risk ⚙️ External commands🌐 Network access🔑 Env variables📁 Filesystem access

Enhance and upscale videos with AI

Process raw or low-quality video files with AI-powered upscaling. This skill automates download, trim, upload, and enhancement for local files and streaming URLs.

Supports: Claude Codex Code(CC)
⚠️ 61 Poor
1

Download the skill ZIP

2

Upload in Claude

Go to Settings → Capabilities → Skills → Upload skill

3

Toggle on and start using

Test it

Using "video-enhancement". /video-enhancement -v ./video.mp4 --hd

Expected outcome:

Job created successfully (ID: 123). Polling status... COMPLETED. Enhanced video available at: https://img.panpan8.com/video-enhance/result_123.mp4

Using "video-enhancement". /video-enhancement -v "https://youtube.com/watch?v=abc" --end 20

Expected outcome:

Downloaded video to /tmp/verging-video-enhancement/input.mp4. Trimmed to 20s. Credits: 60. Job created (ID: 456). Result: https://img.panpan8.com/video-enhance/result_456.mp4

Security Audit

Medium Risk
v4 • 6/8/2026

This skill is a legitimate video enhancement tool that wraps the verging.ai API. Static analysis flagged shell commands (ffmpeg, curl, yt-dlp), env var access (VERGING_API_KEY), network calls to verging.ai, and temp directory usage. After evaluation, all findings are false positives: the shell commands are standard video processing tool invocations documented for instructional purposes, the API key access follows best practices (env var only), and network calls are to the documented verging.ai API. The combination heuristic is expected for a skill that processes media files via a cloud API. No malicious intent detected.

2
Files scanned
215
Lines analyzed
8
findings
4
Total audits
Medium Risk Issues (2)
Shell command execution via ffmpeg, curl, and yt-dlp
The skill instructs execution of shell commands including ffmpeg for video trimming, curl for API calls, and yt-dlp for video downloads. These are standard media processing tools invoked with documented parameters. While the execution pattern triggers static analysis, the commands are purposeful and scoped to the video enhancement workflow.
API key handling via environment variable
The skill accesses VERGING_API_KEY environment variable for authentication with the verging.ai API. This is the recommended secure pattern for credential management. The key is never hardcoded; users set it via export. Privacy section explicitly warns against exposing it in public repos.
Low Risk Issues (2)
Network requests to verging.ai API endpoints
The skill makes HTTP requests to verging.ai/api/v1/* endpoints for authentication, video upload, job creation, and status polling. All requests are to the documented API of the service the skill is designed to interface with. No requests to suspicious or external third-party domains.
Temp directory usage for video processing
The skill uses /tmp/verging-video-enhancement/ as a working directory for trimmed video files. This is a standard ephemeral location for media processing. The skill notes temp files are cleaned up after use.
Audited by: claude View Audit History →

Quality Score

38
Architecture
100
Maintainability
85
Content
23
Community
51
Security
91
Spec Compliance

What You Can Build

Social media content creator

Quickly upscale short clips for Instagram, TikTok, or YouTube Shorts without manual editing software.

Video restoration hobbyist

Enhance old or low-quality home videos to modern resolution standards with one command.

Developer prototyping media tools

Use the verging.ai API workflow as a reference for building video processing pipelines.

Try These Prompts

Enhance a local video file
/video-enhancement -v ./my-video.mp4
Enhance a YouTube video
/video-enhancement -v "https://youtube.com/watch?v=example"
HD upscale with auto-download
/video-enhancement -v ./video.mp4 --hd --download -o ./output/
Trim and enhance a segment
/video-enhancement -v ./long-video.mp4 --start 10 --end 25 --hd

Best Practices

  • Set VERGING_API_KEY as an environment variable rather than passing it inline in commands
  • Trim videos to the required segment before upscaling to minimize credit usage
  • Verify you have rights to process the media before uploading to the API

Avoid

  • Do not hardcode API keys in scripts or commit them to version control
  • Do not process videos longer than 30 seconds without trimming first; the API will reject them
  • Do not use JSON body for upload or job creation endpoints; both require form data with -F flags

Frequently Asked Questions

What tools need to be installed?
You need yt-dlp, ffmpeg, ffprobe, and curl available on your system PATH.
How much does enhancement cost?
Normal mode costs 3 credits per second. HD mode costs 5 credits per second.
Can I process videos longer than 30 seconds?
The API limits jobs to 30 seconds. Use --start and --end flags to trim longer videos first.
Where do enhanced videos get saved?
By default the result URL is displayed. Use --download flag to save to your current directory or specify --output for a custom path.
Is my API key safe?
The key is read from the VERGING_API_KEY environment variable. Never paste it in public repos or shared terminals.
Which video sources are supported?
Local files of any common format plus YouTube and Bilibili URLs through yt-dlp.

Developer Details

File structure

📄 README.md

📄 SKILL.md