Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when the user wants to find YouTube content on any topic: searching for videos or channels, finding creators who cover a subject, discovering tutorials, talks, or expert discussions, or looking up a channel by name or handle. Also use proactively when the user wants to research a topic and YouTube is a good source. Not for account management or written-source-only research.
.claude/skills/zeropointrepo-youtube-search/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 23% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 90% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -11% | 0% |
Search YouTube and fetch transcripts via TranscriptAPI.com.
If $TRANSCRIPT_API_KEY is not set, read references/auth-setup.md and follow the instructions there to get and store the key.
Every request needs two headers:
Bearer $TRANSCRIPT_API_KEYHermesAgent/0.11.0, ClaudeCode/1.0). Version is optional — agent name alone is fine. Do not omit this header or send a bare default — Cloudflare will return a 403 (error code 1010) and block the request.Full OpenAPI spec: transcriptapi.com/openapi.json — consult this for the latest parameters and schemas.
Search YouTube globally for videos or channels.
httpGET https://transcriptapi.com/api/v2/youtube/search?q=QUERY&type=video&limit=20 Authorization: Bearer $TRANSCRIPT_API_KEY User-Agent: YourAgent/1.0
| Param | Required | Default | Validation | | ------- | -------- | ------- | --------------------- | | q | yes | — | 1-200 chars (trimmed) | | type | no | video | video or channel | | limit | no | 20 | 1-50 |
Video search response:
json{ "results": [ { "type": "video", "videoId": "dQw4w9WgXcQ", "title": "Rick Astley - Never Gonna Give You Up", "channelId": "UCuAXFkgsw1L7xaCfnd5JJOw", "channelTitle": "Rick Astley", "channelHandle": "@RickAstley", "channelVerified": true, "lengthText": "3:33", "viewCountText": "1.5B views", "publishedTimeText": "14 years ago", "hasCaptions": true, "thumbnails": [{ "url": "...", "width": 120, "height": 90 }] } ], "result_count": 20 }
Channel search response (type=channel):
json{ "results": [{ "type": "channel", "channelId": "UCuAXFkgsw1L7xaCfnd5JJOw", "title": "Rick Astley", "handle": "@RickAstley", "url": "https://www.youtube.com/@RickAstley", "description": "Official channel...", "subscriberCount": "4.2M subscribers", "verified": true, "rssUrl": "https://www.youtube.com/feeds/videos.xml?channel_id=UC...", "thumbnails": [...] }], "result_count": 5 }
Search videos within a specific channel. Accepts channel — an @handle, channel URL, or UC... ID.
httpGET https://transcriptapi.com/api/v2/youtube/channel/search?channel=@TED&q=climate+change&limit=30 Authorization: Bearer $TRANSCRIPT_API_KEY User-Agent: YourAgent/1.0
| Param | Required | Validation | | --------- | -------- | ----------------------------------------- | | channel | yes | @handle, channel URL, or UC... ID | | q | yes | 1-200 chars | | limit | no | 1-50 (default 30) |
Returns up to ~30 results (YouTube limit). Same video response shape as global search.
Convert @handle to channel ID:
httpGET https://transcriptapi.com/api/v2/youtube/channel/resolve?input=@TED Authorization: Bearer $TRANSCRIPT_API_KEY User-Agent: YourAgent/1.0
http# 1. Search for videos GET https://transcriptapi.com/api/v2/youtube/search?q=python+web+scraping&type=video&limit=5 Authorization: Bearer $TRANSCRIPT_API_KEY User-Agent: YourAgent/1.0 # 2. Get transcript from result GET https://transcriptapi.com/api/v2/youtube/transcript?video_url=VIDEO_ID&format=text&include_timestamp=true&send_metadata=true Authorization: Bearer $TRANSCRIPT_API_KEY User-Agent: YourAgent/1.0
| Code | Meaning | Action | | -------- | ---------------- | ---------------------------------------------- | | 401 | Bad API key | Check key | | 402 | No credits | transcriptapi.com/billing | | 403/1010 | Cloudflare block | Add or fix User-Agent header | | 404 | Not found | Resource doesn't exist | | 408 | Timeout | Retry once | | 422 | Invalid param | Check param format |
Free tier: 100 credits, 300 req/min.
Every request in this file as a ready-to-run one-liner: references/curl-examples.md
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 5,812 | 7,100 | +22% | 1 | 1 | 0% | 799 | 1,723 | +116% | 0 | 0 | — |
case-02 | fail→fail | 12,075 | 5,693 | -53% | 1 | 1 | 0% | 2,093 | 1,716 | -18% | 0 | 0 | — |
case-03 | fail→fail | 8,192 | 8,952 | +9% | 1 | 1 | 0% | 1,419 | 1,970 | +39% | 0 | 0 | — |
case-04 | pass→pass | 12,011 | 4,773 | -60% | 1 | 1 | 0% | 2,010 | 2,293 | +14% | 0 | 0 | — |
case-05 | fail→pass | 11,186 | 2,348 | -79% | 1 | 1 | 0% | 1,867 | 1,791 | -4% | 0 | 0 | — |
case-06 | pass→pass | 8,264 | 2,034 | -75% | 1 | 1 | 0% | 1,208 | 1,717 | +42% | 0 | 0 | — |
case-07 | pass→pass | 9,795 | 3,089 | -68% | 1 | 1 | 0% | 1,545 | 1,843 | +19% | 0 | 0 | — |
case-08 | fail→pass | 10,292 | 1,785 | -83% | 1 | 1 | 0% | 1,334 | 1,645 | +23% | 0 | 0 | — |
case-09 | pass→pass | 9,026 | 3,208 | -64% | 1 | 1 | 0% | 1,584 | 1,971 | +24% | 0 | 0 | — |
case-10 | fail→pass | 10,923 | 4,368 | -60% | 1 | 1 | 0% | 1,873 | 2,054 | +10% | 0 | 0 | — |
case-11 | fail→pass | 6,291 | 2,904 | -54% | 1 | 1 | 0% | 1,021 | 1,935 | +90% | 0 | 0 | — |
case-12 | pass→pass | 6,107 | 1,937 | -68% | 1 | 1 | 0% | 782 | 1,664 | +113% | 0 | 0 | — |
case-13 | pass→pass | 6,286 | 1,858 | -70% | 1 | 1 | 0% | 888 | 1,628 | +83% | 0 | 0 | — |
case-14 | fail→pass | 12,131 | 1,605 | -87% | 1 | 1 | 0% | 1,799 | 1,594 | -11% | 0 | 0 | — |
case-15 | pass→fail | 5,117 | 3,612 | -29% | 1 | 1 | 0% | 768 | 1,959 | +155% | 0 | 0 | — |
case-16 | pass→pass | 8,684 | 2,566 | -70% | 1 | 1 | 0% | 1,393 | 1,843 | +32% | 0 | 0 | — |
case-17 | fail→pass | 5,547 | 2,157 | -61% | 1 | 1 | 0% | 711 | 1,631 | +129% | 0 | 0 | — |
case-18 | pass→pass | 10,083 | 2,213 | -78% | 1 | 1 | 0% | 1,679 | 1,754 | +4% | 0 | 0 | — |
case-19 | fail→pass | 7,825 | 1,978 | -75% | 1 | 1 | 0% | 1,196 | 1,656 | +38% | 0 | 0 | — |
case-20 | pass→pass | 6,462 | 4,538 | -30% | 1 | 1 | 0% | 990 | 2,103 | +112% | 0 | 0 | — |
case-21 | pass→pass | 10,182 | 5,070 | -50% | 1 | 1 | 0% | 1,692 | 2,313 | +37% | 0 | 0 | — |
case-22 | pass→pass | 9,803 | 3,715 | -62% | 1 | 1 | 0% | 1,737 | 1,907 | +10% | 0 | 0 | — |
case-23 | fail→pass | 9,284 | 2,028 | -78% | 1 | 1 | 0% | 1,371 | 1,631 | +19% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted, and 20 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +30 percentage points is the difference between those two pass rates over the 20 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.