worthclip-youtube-video-scorer
AI-powered YouTube video scoring.
Setup & Installation
Install command
clawhub install ivanstancich/worthclip-youtube-video-scorerIf the CLI is not installed:
Install command
npx clawhub@latest install ivanstancich/worthclip-youtube-video-scorerOr install with OpenClaw CLI:
Install command
openclaw skills install ivanstancich/worthclip-youtube-video-scoreror paste the repo link into your assistant's chat
Install command
https://github.com/openclaw/skills/tree/main/skills/ivanstancich/worthclip-youtube-video-scorerWhat This Skill Does
WorthClip scores YouTube videos 1-10 against a personalized learning profile. It returns AI summaries, alignment analysis, and a curated feed of scored content. Supports channel tracking, persona configuration, and paginated feed browsing.
Instead of manually evaluating whether a video matches your learning goals, WorthClip automates scoring so you can triage a backlog of videos quickly.
When to Use It
- Score a YouTube video before watching it
- Filter your feed to only show videos scored 8 or above
- Track specific YouTube channels and score their uploads
- Update your learning goals to recalibrate video scoring
- Check API usage to stay within billing limits
View original SKILL.md file
# WorthClip - YouTube Video Scoring
Score YouTube videos 1-10 based on your personalized learning goals. Get AI-powered summaries, alignment analysis, and a curated feed.
## Setup
1. Sign up at https://worthclip.com
2. Go to Settings > API Keys
3. Generate an API key
4. Set it: `export WORTHCLIP_API_KEY="wc_your_key_here"`
## Commands
### Score a video
Scores a YouTube video against the user's persona and goals. Handles async scoring automatically with polling.
```bash
bash {baseDir}/scripts/score.sh "VIDEO_ID"
```
The script submits the video for scoring, polls for completion (up to 60 seconds), and returns the completed score JSON. If the video was already scored, it returns the existing score immediately.
### Get your feed
Returns scored videos sorted by relevance, with optional filters.
```bash
bash {baseDir}/scripts/feed.sh [--min-score N] [--verdict VERDICT] [--limit N] [--cursor N]
```
Options:
- `--min-score N` - Only return videos scored N or above (1-10)
- `--verdict VERDICT` - Filter by verdict (e.g., "watch", "skip")
- `--limit N` - Number of results per page
- `--cursor N` - Pagination cursor from previous response
### Check usage
Shows current billing period usage stats and limits.
```bash
bash {baseDir}/scripts/usage.sh
```
## API Reference
Base URL: `https://greedy-mallard-11.convex.site/api/v1`
The API is hosted on Convex (convex.site), WorthClip's serverless backend. The domain `greedy-mallard-11.convex.site` is WorthClip's production Convex deployment. You can verify this by visiting https://worthclip.com/developers.
All requests (except /health) require `Authorization: Bearer YOUR_API_KEY` header.
| Endpoint | Method | Description |
|----------|--------|-------------|
| /health | GET | Health check (no auth required) |
| /score | POST | Score a video (async, returns 202 with jobId) |
| /score/:jobId | GET | Poll scoring job status |
| /videos/:ytId/summary | GET | Get video summary (summarization) |
| /videos/:ytId | GET | Get video detail with full score |
| /feed | GET | Paginated scored feed with filters |
| /channels | GET | List tracked channels |
| /channels/lookup | POST | Lookup channel by YouTube URL |
| /channels/track | POST | Track a new channel |
| /persona | GET | Get current persona and goals |
| /persona | PUT | Update persona description |
| /goals | PUT | Update learning goals |
| /usage | GET | Current billing period usage stats |
## Rate Limits
- **General:** 60 requests/minute (all endpoints)
- **Scoring:** 20 requests/minute (POST /score and GET /score/:jobId)
Response headers:
- `X-RateLimit-Limit` - Maximum requests per window
- `X-RateLimit-Remaining` - Requests remaining in current window
- `Retry-After` - Seconds to wait before retrying (only on 429 responses)
## Error Format
All errors return a consistent JSON structure with an appropriate HTTP status code:
```json
{
"error": {
"code": "ERROR_CODE",
"message": "Human-readable description of the error"
}
}
```
Common error codes:
- `UNAUTHORIZED` (401) - Missing or invalid API key
- `RATE_LIMITED` (429) - Too many requests
- `NOT_FOUND` (404) - Resource not found
- `VALIDATION_ERROR` (400) - Invalid request parameters
- `INTERNAL_ERROR` (500) - Server error
## Base URL
```
https://greedy-mallard-11.convex.site/api/v1
```
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Score a YouTube video before watching it
- 1Score a YouTube video before watching it
- 2Filter your feed to only show videos scored 8 or above
- 3Track specific YouTube channels and score their uploads
- 4Update your learning goals to recalibrate video scoring
- 5Check API usage to stay within billing limits
AI-powered YouTube video scoring.
Security Audits
These signals reflect official OpenClaw status values. A Suspicious status means the skill should be used with extra caution.