agentleFS
baofeng-techbaofeng-tech/agent-skills-so/youtube-search/SKILL.md

youtube-search

Skill0 starsChanged 7 months agoMITView on GitHub

YouTube Search API via AIsa unified endpoint. Search YouTube videos, channels, and playlists with a single AIsa API key — no Google API key or OAuth required. Use this skill when users want to search YouTube content. For other AIsa capabilities (LLM, financial data, Twitter, web search), see the aisa-core skill. Use when: the user needs YouTube search, trend discovery, channel research, or SERP analysis.

  • Reads credentials

13 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

Used with

Projects with it also use youtube-transcript.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • dandacompanySkill

    youtube-transcript

    Extract YouTube video transcripts/subtitles using yt-dlp. Handles rate limiting, auto-generated captions, multiple languages, and VTT parsing.

    Also in 4 projects that use this

The file

We split this skill into the rules it gives an agent

12 rules. 0 are common, 6 are distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 12 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Search YouTube videos, channels, and playlists through AIsa's unified API.

    line 18
  2. Distinctive:

    All YouTube search requests go through this single endpoint.

    line 41
  3. Distinctive:

    All requests require the AIsa API key in the Authorization header:

    line 45
  4. Very specific:

    Use the next_page_token from a response to fetch the next page of results:

    line 204
  5. Distinctive:

    All requests are pay-per-use through your AIsa balance — no separate YouTube API quota management

    line 359
  6. Very specific:

    Use next_page_token from previous responses as the sp value for pagination

    line 363
  7. Very specific:

    Check that the AISA_API_KEY environment variable is set:

    line 382
  8. Distinctive:

    Verify the API key works with a simple test:

    line 388
  9. Distinctive:

    Verify your query is not empty

    line 408
  10. Very specific:

    If using gl, verify the country code is supported — not all ISO codes work (e.g., cn is unsupported).

    line 410
  11. Very specific:

    Ensure engine=youtube is included in every request

    line 411
  12. Very specific:

    Check if results are in sections instead of videos (common for non-English queries)

    line 412

“Common” is measured: kept by at least 6 in 10 of the 12 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

YouTube Search (via AIsa)

Search YouTube videos, channels, and playlists through AIsa's unified API. No Google API key or OAuth setup needed — just your AIsa API key.

Quick Start

# Search for videos (using requests — recommended)
python <<'EOF'
import os, json, requests
results = requests.get(
    'https://api.aisa.one/apis/v1/youtube/search',
    headers={'Authorization': f'Bearer {os.environ["AISA_API_KEY"]}'},
    params={'engine': 'youtube', 'q': 'coding tutorial'}
).json()
print(json.dumps(results, indent=2))
EOF

Base URL

https://api.aisa.one/apis/v1/youtube/search

All YouTube search requests go through this single endpoint. AIsa handles authentication with the underlying YouTube data source — you only need your AIsa API key.

Authentication

All requests require the AIsa API key in the Authorization header:

Authorization: Bearer $AISA_API_KEY

Environment Variable: Set your API key as AISA_API_KEY:

export AISA_API_KEY="YOUR_AISA_API_KEY"

Getting Your API Key

  1. Sign in or create an account at AIsa Marketplace
  2. Navigate to your Dashboard
  3. Copy your API key

API Reference

YouTube Search

GET /apis/v1/youtube/search

Query Parameters

ParameterTypeRequiredDescription
enginestringYesMust be youtube
qstringYesSearch query (same syntax as YouTube search box)
spstringNoYouTube filter token for pagination or advanced filters
glstringNoCountry code for localized results (e.g., us, jp, gb). Not all country codes are supported — see notes below
hlstringNoInterface language (e.g., en, zh, ja)

Example: Basic Search

curl -s -X GET "https://api.aisa.one/apis/v1/youtube/search?engine=youtube&q=machine+learning+tutorial" \
  -H "Authorization: Bearer $AISA_API_KEY"

Example: Search with Country & Language

curl -s -X GET "https://api.aisa.one/apis/v1/youtube/search?engine=youtube&q=AI+news&gl=us&hl=en" \
  -H "Authorization: Bearer $AISA_API_KEY"

Example: Pagination with sp Token

# Use the sp token from a previous response to get the next page
curl -s -X GET "https://api.aisa.one/apis/v1/youtube/search?engine=youtube&q=python+tutorial&sp=EgIQAQ%3D%3D" \
  -H "Authorization: Bearer $AISA_API_KEY"

Response

The API returns structured YouTube search results including video metadata, channel info, thumbnails, and pagination tokens.

Note: The response structure may vary by query language. English queries typically return results in the videos array. Some non-English queries may return results grouped in a sections array instead. Always check for both fields.

{
  "search_metadata": {
    "status": "Success",
    "total_time_taken": 1.2
  },
  "search_parameters": {
    "engine": "youtube",
    "q": "machine learning tutorial"
  },
  "next_page_token": "CBQQABoCEgA%3D",
  "videos": [
    {
      "position_on_page": 1,
      "title": "Machine Learning Full Course for Beginners",
      "link": "https://www.youtube.com/watch?v=abc123xyz",
      "channel": {
        "name": "Tech Academy",
        "link": "https://www.youtube.com/channel/UCxyz123",
        "thumbnail": "https://yt3.ggpht.com/..."
      },
      "published_date": "2 months ago",
      "views": 1500000,
      "length": "3:45:20",
      "description": "Complete machine learning tutorial...",
      "thumbnail": {
        "static": "https://i.ytimg.com/vi/abc123xyz/hq720.jpg",
        "rich": "https://i.ytimg.com/an_webp/abc123xyz/mqdefault_6s.webp"
      }
    }
  ]
}

Alternate response structure (non-English / some queries):

Some queries return results grouped in sections instead of a flat videos array:

{
  "sections": [
    {
      "title": "搜索结果",
      "videos": [
        {
          "title": "编程教程...",
          "link": "https://www.youtube.com/watch?v=...",
          ...
        }
      ]
    }
  ]
}

Parsing both formats:

# Handle both response structures
videos = results.get('videos', [])
if not videos and 'sections' in results:
    for section in results['sections']:
        videos.extend(section.get('videos', []))

Advanced Search Tips

YouTube's q parameter supports the same search syntax as the YouTube search box:

Search SyntaxDescriptionExample
Basic keywordsStandard searchq=python tutorial
Exact phraseQuote for exact matchq="machine learning basics"
Channel filterSearch within a channelq=channel:GoogleDevelopers python
Duration hintCombine with keywordsq=python tutorial long

Using the sp Filter Token

The sp parameter accepts YouTube's encoded filter tokens. Common values:

Filtersp ValueDescription
Videos onlyEgIQAQ%3D%3DFilter to video results only
Channels onlyEgIQAg%3D%3DFilter to channel results only
Playlists onlyEgIQAw%3D%3DFilter to playlist results only
Live nowEgJAAQ%3D%3DCurrently live streams
This weekEgIIAw%3D%3DUploaded this week
This monthEgIIBA%3D%3DUploaded this month
Short (<4 min)EgIYAQ%3D%3DShort duration videos
Long (>20 min)EgIYAg%3D%3DLong duration videos

You can also obtain sp tokens from the next_page_token field in previous API responses for pagination.

Pagination

Use the next_page_token from a response to fetch the next page of results:

# First page
results = requests.get(
    'https://api.aisa.one/apis/v1/youtube/search',
    headers=headers,
    params={'engine': 'youtube', 'q': 'python tutorial'}
).json()

# Get next page token
next_token = results.get('next_page_token')
if next_token:
    page2 = requests.get(
        'https://api.aisa.one/apis/v1/youtube/search',
        headers=headers,
        params={'engine': 'youtube', 'q': 'python tutorial', 'sp': next_token}
    ).json()

Code Examples

JavaScript

const headers = {
  'Authorization': `Bearer ${process.env.AISA_API_KEY}`
};

// Basic YouTube search
const results = await fetch(
  'https://api.aisa.one/apis/v1/youtube/search?engine=youtube&q=AI+agents+tutorial',
  { headers }
).then(r => r.json());

console.log(results.videos);

// Search with filters
const filtered = await fetch(
  'https://api.aisa.one/apis/v1/youtube/search?engine=youtube&q=deep+learning&gl=us&hl=en&sp=EgIQAQ%3D%3D',
  { headers }
).then(r => r.json());

Python

import os
import requests

headers = {'Authorization': f'Bearer {os.environ["AISA_API_KEY"]}'}

# Basic YouTube search
results = requests.get(
    'https://api.aisa.one/apis/v1/youtube/search',
    headers=headers,
    params={'engine': 'youtube', 'q': 'AI agents tutorial'}
).json()

for video in results.get('videos', []):
    print(f"{video['title']} - {video.get('views', 'N/A')} views")

# Search with country and language
results_jp = requests.get(
    'https://api.aisa.one/apis/v1/youtube/search',
    headers=headers,
    params={'engine': 'youtube', 'q': 'プログラミング', 'gl': 'jp', 'hl': 'ja'}
).json()

Python (urllib, no dependencies)

Note: urllib may encounter 403 errors due to its default User-Agent. Using requests (above) is recommended. If you must use urllib, always set a custom User-Agent header.

import urllib.request, urllib.parse, os, json

def youtube_search(query, gl=None, hl=None, sp=None):
    """Search YouTube via AIsa API."""
    params = {'engine': 'youtube', 'q': query}
    if gl: params['gl'] = gl
    if hl: params['hl'] = hl
    if sp: params['sp'] = sp
    
    url = f'https://api.aisa.one/apis/v1/youtube/search?{urllib.parse.urlencode(params)}'
    req = urllib.request.Request(url)
    req.add_header('Authorization', f'Bearer {os.environ["AISA_API_KEY"]}')
    req.add_header('User-Agent', 'AIsa-Skill/1.0')
    return json.load(urllib.request.urlopen(req))

# Search
results = youtube_search('OpenClaw tutorial', gl='us', hl='en')

# Handle both response formats
videos = results.get('videos', [])
if not videos and 'sections' in results:
    for section in results['sections']:
        videos.extend(section.get('videos', []))

print(json.dumps(videos[:3], indent=2))

Combining with Other AIsa APIs

One of the key advantages of AIsa is the unified API key. Use the same AISA_API_KEY to combine YouTube search with other AIsa capabilities:

YouTube Search + LLM Summary

import os, requests, json

headers = {'Authorization': f'Bearer {os.environ["AISA_API_KEY"]}'}

# 1. Search YouTube
yt_results = requests.get(
    'https://api.aisa.one/apis/v1/youtube/search',
    headers=headers,
    params={'engine': 'youtube', 'q': 'latest AI developments 2026'}
).json()

# 2. Summarize with LLM (same API key!)
video_titles = [v['title'] for v in yt_results.get('videos', [])[:5]]
summary = requests.post(
    'https://api.aisa.one/v1/chat/completions',
    headers={**headers, 'Content-Type': 'application/json'},
    json={
        'model': 'qwen3-flash',
        'messages': [
            {'role': 'user', 'content': f'Summarize the trending AI topics based on these YouTube videos: {json.dumps(video_titles)}'}
        ]
    }
).json()

print(summary['choices'][0]['message']['content'])

YouTube Search + Web Search

# Search both YouTube and the web for comprehensive research
yt_results = requests.get(
    'https://api.aisa.one/apis/v1/youtube/search',
    headers=headers,
    params={'engine': 'youtube', 'q': 'AI agent frameworks 2026'}
).json()

web_results = requests.get(
    'https://api.aisa.one/apis/v1/search/smart',
    headers=headers,
    params={'q': 'AI agent frameworks 2026'}
).json()

Notes

  • All requests are pay-per-use through your AIsa balance — no separate YouTube API quota management
  • The engine parameter must always be set to youtube
  • Video URLs follow the format https://www.youtube.com/watch?v={videoId}
  • Channel URLs follow the format https://www.youtube.com/channel/{channelId}
  • Use next_page_token from previous responses as the sp value for pagination
  • The gl (country) parameter does not support all ISO country codes. Known unsupported values include cn (China). If you get Unsupported value errors, try omitting gl or use a different country code
  • Non-English queries may return results in a sections array instead of a flat videos array — always handle both formats
  • IMPORTANT: Python urllib may return 403 errors due to its default User-Agent. Use the requests library instead, or add a custom User-Agent header
  • IMPORTANT: When using curl commands, ensure environment variables like $AISA_API_KEY are properly expanded
  • IMPORTANT: When piping curl output to jq, use -s flag and ensure the API key is set

Error Handling

StatusMeaning
200Successful search response
400Invalid request parameters (missing engine or q)
401Unauthorized — invalid or missing AIsa API key
429Rate limited
500Internal server error

Troubleshooting: API Key Issues

  1. Check that the AISA_API_KEY environment variable is set:
echo $AISA_API_KEY
  1. Verify the API key works with a simple test:
python <<'EOF'
import os, json, requests
try:
    result = requests.get(
        'https://api.aisa.one/apis/v1/youtube/search',
        headers={'Authorization': f'Bearer {os.environ["AISA_API_KEY"]}'},
        params={'engine': 'youtube', 'q': 'test'}
    ).json()
    videos = result.get('videos', [])
    print(f"✅ API key is valid. Results: {len(videos)} videos found")
except Exception as e:
    print(f"❌ Error: {e}")
EOF

Troubleshooting: No Results

  1. Verify your query is not empty
  2. Try a broader search term
  3. If using gl, verify the country code is supported — not all ISO codes work (e.g., cn is unsupported). Try omitting gl to test
  4. Ensure engine=youtube is included in every request
  5. Check if results are in sections instead of videos (common for non-English queries)

Resources

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.