iopho-searching-videos — Cross-Platform Video Search

Search for videos across multiple platforms and return structured metadata — without downloading anything.

Prerequisites

Check what's available:

!which yt-dlp 2>/dev/null && echo "yt-dlp: ✓" || echo "yt-dlp: ✗ — install via: pip install yt-dlp or brew install yt-dlp"

!python3 -c "from duckduckgo_search import DDGS; print('duckduckgo-search: ✓')" 2>/dev/null || echo "duckduckgo-search: ✗ (optional) — install via: pip install duckduckgo-search"

Minimum requirement: yt-dlp (covers 19+ platforms with native search). Optional: duckduckgo-search for cross-platform web video search aggregation.

Platform Search Matrix

Quick Reference

Basic search (YouTube, top 5):

yt-dlp --flat-playlist --dump-json "ytsearch5:$QUERY" 2>/dev/null | python3 -c "
import sys, json
for line in sys.stdin:
    v = json.loads(line)
    dur = f\"{v.get('duration',0)//60}:{v.get('duration',0)%60:02d}\" if v.get('duration') else '?'
    print(f\"  {v.get('title','?')}\")
    print(f\"    Duration: {dur} | Views: {v.get('view_count','?')} | URL: https://youtube.com/watch?v={v['id']}\")
    print()
"

Bilibili search:

yt-dlp --flat-playlist --dump-json "BiliBiliSearch:$QUERY" 2>/dev/null | python3 -c "
import sys, json
for line in sys.stdin:
    v = json.loads(line)
    dur = f\"{v.get('duration',0)//60}:{v.get('duration',0)%60:02d}\" if v.get('duration') else '?'
    print(f\"  {v.get('title','?')}\")
    print(f\"    Duration: {dur} | URL: {v.get('url', v.get('webpage_url', '?'))}\")
    print()
"

Filter by duration (e.g., under 30 seconds):

yt-dlp --flat-playlist --dump-json --match-filter "duration<30" "ytsearch20:$QUERY"

Full JSON metadata (for programmatic use):

yt-dlp --flat-playlist --dump-json "ytsearch5:$QUERY" 2>/dev/null

Web-wide video search (via DuckDuckGo, no API key needed):

python3 -c "
from duckduckgo_search import DDGS
results = DDGS().videos('$QUERY', max_results=5)
for v in results:
    print(f\"  {v.get('title','?')}\")
    print(f\"    Duration: {v.get('duration','?')} | Source: {v.get('publisher','?')} | URL: {v.get('content', '?')}\")
    print()
"

Execute

Parse the user's arguments and run the appropriate search:

# Parse arguments from: $ARGUMENTS
# Expected format: <query> [--platform youtube|bilibili|all] [--limit N] [--json]
#
# Default: --platform youtube --limit 5
#
# The agent should:
# 1. Extract the query, platform, limit, and json flag from $ARGUMENTS
# 2. Choose the right search prefix based on platform
# 3. Run yt-dlp with --flat-playlist --dump-json
# 4. Format output as a readable table (or raw JSON if --json)
# 5. If --platform all, also try DuckDuckGo if available

Advanced Usage

Search with date sorting (YouTube only):

yt-dlp --flat-playlist --dump-json "ytsearchdate10:query"

Search and preview metadata of a specific video:

yt-dlp --dump-json "https://youtube.com/watch?v=VIDEO_ID" 2>/dev/null | python3 -c "
import sys, json
v = json.loads(sys.stdin.read())
print(f\"Title: {v['title']}\")
print(f\"Duration: {v['duration']}s | Views: {v.get('view_count','?')}\")
print(f\"Channel: {v.get('channel','?')} | Upload: {v.get('upload_date','?')}\")
print(f\"Description: {v.get('description','')[:200]}...\")
print(f\"Tags: {', '.join(v.get('tags',[])) if v.get('tags') else 'none'}\")
"

Combine with iopho-getting-videos: After finding videos, use /iopho-getting-videos <url> to download.

Supported yt-dlp Search Extractors

Full list of platforms with built-in search support: YouTube, SoundCloud, Bilibili, Niconico, Dailymotion, Google Video, Yahoo, GameJolt, Mail.ru, Rokfin, and more.

Run yt-dlp --list-extractors | grep -i search to see all available search extractors.

Tips

• Use --match-filter "duration<60" to find short videos (< 1 minute)
• Use --match-filter "view_count>10000" to find popular videos
• Combine filters: --match-filter "duration<30 & view_count>1000"
• For B站, search queries work best in Chinese
• DuckDuckGo search covers platforms that yt-dlp doesn't have native search for (Vimeo, TikTok, Twitter/X, etc.)