Films Search Skill

Prerequisites

• web-search skill (required, used to search and discover resource pages)

Commands

Search for Resources

bash "$SKILLS_ROOT/films-search/scripts/film-search.sh" search "keyword" [options]

--pan <type>

quark

baidu

aliyun

uc

all

all

--quality <q>

4k

1080p

720p

all

all

--limit <n>

5

--engine <e>

deep

web

deep

• deep

  (default, recommended) — Use web-search skill to find resource pages + JavaScript deep crawling to extract cloud drive links and access codes, with the most accurate results

• web

  — Only extract links from web-search engine summaries (fast, but lower accuracy, no deep crawling)

# Search resources on all platforms (uses deep search by default)
bash "$SKILLS_ROOT/films-search/scripts/film-search.sh" search "The Wandering Earth 2"

# Only search for 4K resources on Quark cloud drive
bash "$SKILLS_ROOT/films-search/scripts/film-search.sh" search "The Wandering Earth 2" --pan quark --quality 4k

# Limit the number of results
bash "$SKILLS_ROOT/films-search/scripts/film-search.sh" search "The Wandering Earth 2" --limit 10

# Use shallow search (no deep page crawling)
bash "$SKILLS_ROOT/films-search/scripts/film-search.sh" search "The Wandering Earth 2" --engine web

powershell -File "$SKILLS_ROOT/films-search/scripts/film-search.ps1" search "The Wandering Earth 2" --pan quark

Hot Recommendations

bash "$SKILLS_ROOT/films-search/scripts/film-search.sh" hot "2025 Popular Movies"

Resolve Redirect Links

bash "$SKILLS_ROOT/films-search/scripts/film-search.sh" resolve "https://example.com/goto/xxx"

Agent Usage Flow

1. User says: "Help me find the movie XXX" → Execute

   search "XXX"

2. Parse the returned JSON and extract cloud drive links
3. Present to the user grouped by cloud drive type and video quality
4. If there is an access code (

   extractCode

   ) in the results, display it together
5. If there is a

   pageUrl

   but no direct

   url

   , use the

   resolve

   command to get the real address

User: Help me find 4K Quark cloud drive resources for The Wandering Earth 2

Agent:

1. Execute

   search "The Wandering Earth 2" --pan quark --quality 4k

2. Extract matching items from the JSON results
3. Return: Title + Video Quality + Quark Cloud Drive Link + Access Code (if any)

Output Format

{
  "success": true,
  "data": {
    "query": "The Wandering Earth 2",
    "total": 5,
    "results": [
      {
        "title": "Resource Title",
        "pan": "quark",
        "url": "https://pan.quark.cn/s/xxx",
        "quality": "4K",
        "extractCode": "ab12",
        "source": "deep-search",
        "pageUrl": "https://example.com/resource/123"
      }
    ]
  }
}

• pan

  : Cloud drive type —

  quark

  (Quark),

  baidu

  (Baidu),

  aliyun

  (Alibaba),

  uc

  (UC),

  magnet

  (Magnet Link)

• quality

  : Detected video quality —

  4K

  ,

  1080P

  ,

  720P

  , empty string if not detected

• source

  : Result source —

  deep-search

  (Deep Crawling),

  web-search

  (Search Summary)

• url

  : Directly accessible cloud drive sharing link

• extractCode

  : Access code/password (if detected)

• pageUrl

  : URL of the source page for the result

Configuration

films-search/.env

# Preferred cloud drive types (comma-separated, earlier ones are prioritized)
FILM_SEARCH_PREFERRED_PAN=quark,aliyun,baidu,uc

# Deep search toggle (recommended to enable)
FILM_SEARCH_DEEP_ENABLED=true

# Maximum number of result pages accessed per search
FILM_SEARCH_DEEP_MAX_PAGES=6

# Number of concurrent page crawls
FILM_SEARCH_DEEP_CONCURRENCY=4

Notes

• It is recommended to use Chinese film titles for the best search results
• Deep search (

  --engine deep

  ) takes 2-3 seconds longer than shallow search, but results are much more accurate
• Deep search requires the web-search skill to be enabled; the first search may require waiting for the browser to start
• Cloud drive sharing links may become invalid at any time, so it is recommended to use them as soon as possible

Disclaimer