ae_instagram_fetch_search
Search Instagram for users, hashtags, or places by keyword.
Inputs:
- `query` (required): The search keyword.
- `select` (optional): Restrict results to `users`, `hashtags`, or `places`. Omit it to search all supported types.
Returns the matching results in the response data, along with response status and request metadata.
ae_xiaohongshu_pgy_creator_get_profile
通过已知的小红书博主用户ID,获取该创作者在蒲公英平台的主页资料及合作报价信息。参数请放在必填的 request_body 对象中。
ae_douyin_billboard_fetch_hot_total_hot_word_list
## Purpose
Retrieve a page from Douyin's all-content hot-word list by search keyword and recent time window.
## Inputs
All inputs are optional and belong inside the `request_body` object:
- `keyword` (string): Search keyword. Default: `"抖音"`.
- `page_num` (integer): Page number. Default: `1`.
- `page_size` (integer): Number of items per page. Default: `10`.
- `date_window` (integer): Time window in hours. Documented choices are `24`, `72`, and `168`, meaning the last 1, 3, and 7 days respectively. Default: `24`.
## Example
```json
{
"request_body": {
"keyword": "抖音",
"page_num": 1,
"page_size": 10,
"date_window": 24
}
}
```
## Result
Returns a response object with content-word results in `data`, which may be null. The response schema does not specify the fields of individual results or pagination metadata. A single request retrieves a page, not necessarily the entire matching list.
ae_douyin_index_fetch_multi_keyword_hot_trend
Retrieve daily popularity trends for one or more keywords on Douyin or Toutiao to compare how their popularity changes over a specified date range.
Inputs:
- keyword_list (required): Keywords as a comma-separated string, for example "美食,旅游".
- start_date and end_date (required): Start and end dates as YYYYMMDD strings.
- app_name (optional): "aweme" for Douyin (default) or "toutiao" for Toutiao.
- region (optional): Region names as a comma-separated string, for example "云南,上海,北京". Leave empty for nationwide results.
Returns daily keyword popularity trend data for the selected period, platform, and region filter. The documentation does not specify the detailed trend-data fields or measurement units; interpret the values according to the returned data.
ae_tiktok_search_content
Search for comprehensive results matching a keyword.
Parameters:
- `keyword` (required): Search term.
- `offset` (optional, default 0): Starting position for pagination.
- `count` (optional, default 20): Number of results requested.
- `sort_type` (optional, default 0): 0 for relevance; 1 for most likes.
- `publish_time` (optional, default 0): Publication-time filter in days: 0 for no limit, or 1, 7, 30, 90, or 180 days.
Returns the search response, including its result data.
ae_xiaohongshu_get_keyword_related
Find related Xiaohongshu search terms for keyword expansion and content planning.
Input: Provide the required request_body object with search_word, a string of 1–100 characters. No additional fields are accepted in request_body.
Example: {"request_body":{"search_word":"面膜"}}
Results: The business data is at resp["data"]["data"] when present. upSearchWords lists terms searched before the seed keyword; downSearchWords lists terms searched afterward. Each term includes searchScore (search index, not an absolute search-volume count), readTotalCount (total reads), noteNum (note count), relevanceRank (relevance rank), and brandTag (whether the term is a brand keyword). These metrics support comparison of related terms; they do not guarantee low competition. The business data may be null when no result is available.
ae_youtube_playlistitems_list
List the videos and other items in a YouTube playlist.
ae_dumplingai_list_tiktok_video_comments
Fetch comments and replies for a specific TikTok video using DumplingAI.
ae_douyin_fetch_video_comment_replies
## Purpose
Retrieve replies to a specific comment on a Douyin video.
## Inputs
- `item_id` (required, string): ID of the video containing the comment.
- `comment_id` (required, string): ID of the comment whose replies you want to retrieve.
- `cursor` (optional, integer; default `0`): Pagination cursor.
- `count` (optional, integer; default `20`): Requested number of replies.
## Result
Returns a JSON response containing comment-reply data in `data`, together with response status and metadata. The response schema does not define the individual reply fields or pagination metadata, so do not assume a fixed reply structure or that every request returns the requested number of replies. This tool retrieves replies to the selected comment, not all top-level comments on the video.
ae_wechat_get_user_videos
## Purpose
Retrieve a page of videos posted on a specified WeChat Channels account, with video metadata, engagement counts, media links and pagination information. This tool returns data and links; it does not download or decrypt video files.
## Inputs
Supply a `request_body` object containing:
- `username` (required): the account's finder username, in `v2_…@finder` format with hexadecimal characters between the prefix and suffix; 10–256 characters. Use the finder username, not the account's display name or a video ID.
- `last_buffer` (optional; default `""`): a base64 pagination cursor. Omit it or leave it empty for the first page. For subsequent pages, pass the previous response's cursor unchanged when the continuation flag indicates more results.
- `raw` (optional; default `true`): `true` returns the raw structure; `false` returns a simplified structure recommended for media processing.
## Results
With `raw=false`, `data` contains `username`, `nickname`, `count` (the number of videos on this page), `up_continue`, `last_buffer` and `videos[]`. Each video includes identifiers, title, account information, view/like/favorite/share/comment counts, publish time, location and a `media` object. The media fields are `url`, `url_token`, `full_url` (the combined download URL) and `decode_key`.
With `raw=true`, the video array is `data.object[]` (singular `object`). Pagination uses `data.upContinueFlag` and `data.lastBuffer`, with `data.last_buffer` also supplied as an alias. Account information appears in `data.contact` and `data.finderUserInfo`. Media is nested at `data.object[N].objectDesc.media[0]`, with `url`, `urlToken` and `decodeKey`.
## Important notes
- Each call retrieves one page, not necessarily all videos. Continue only when the returned continuation flag indicates another page.
- Preserve `id`, `object_nonce_id` and other 64-bit identifiers as exact strings through lossless JSON parsing. JavaScript `Number` can corrupt their precision.
- Use `full_url`, or concatenate the media URL and its token, for downloading. A successful download does not guarantee a playable file: encrypted video requires separate decryption. Links and keys change on each request, so use the key and encrypted file from the same response.
- Responses may be slow; configure the client request timeout to 30 seconds. This is a client setting, not an additional tool input.
ae_facebook_list_profile_posts
Browse public posts from a Facebook profile using its unique profileId. The optional cursor defaults to an empty string for the initial request; use a pagination cursor from a previous result, if available, to request the next page. Returns a JSON response containing the returned post data in data, along with code and message for the request outcome. Individual post fields and pagination details are not fixed by the response schema, and one request does not guarantee the profile's complete posting history.
ae_douyin_index_fetch_hot_detail
Retrieve analytics for a single Douyin hot topic in one request.
Input:
- topic_name (required string): The hot-topic name. You can use a name obtained from current hot-topic results.
Documented result contents:
- trend_item: Hot-topic index values over time.
- content_item: Popular videos, including titles, links, covers, authors, and like/comment counts.
- play_rank_author_item, digg_rank_author_item, fans_rank_author_item, comment_rank_author_item: Related creators ranked by plays, likes, followers, and comments respectively.
- portrait: Audience distributions by province, city, gender, and age. Values represent proportions in decimal form.
The response contains a data payload that may be null. The analytics fields above are documented contents, not a fully specified response schema. Use this tool for a known topic, not to discover a list of current hot topics.
ae_x_twitter_getpostscountsall
Count matching posts over a specified historical range, grouped by minute, hour or day.
ae_douyin_search_fetch_general_search_v1
## Purpose
Search the Douyin app's general search tab by keyword. This is general search, not the standalone video-search tab. You can sort results and filter by publication time, video duration and content type.
## Inputs
Pass a required `request_body` object. Its fields have documented defaults:
- `keyword`: search text; default `"猫咪"`.
- `cursor`: integer pagination cursor; default `0`.
- `search_id` and `backtrace`: pagination strings; default `""`.
- `sort_type`: `"0"` comprehensive (default), `"1"` most likes, `"2"` latest.
- `publish_time`: `"0"` unrestricted (default), `"1"` last day, `"7"` last week, `"180"` last half year.
- `filter_duration`: `"0"` unrestricted (default), `"0-1"` within one minute, `"1-5"` one to five minutes, `"5-10000"` more than five minutes.
- `content_type`: `"0"` all (default), `"1"` video, `"2"` picture, `"3"` article.
Use strings for all filter values. On the first request, use `cursor: 0`, `search_id: ""` and `backtrace: ""`. For later pages, copy these three values from the previous response.
## Results
The response's `data` payload contains search results. Depending on the item, results may include content IDs and descriptions, author profiles, music, hashtags, video playback and cover URLs, share links, engagement statistics and publishing status or permissions. Common documented fields include `aweme_info`, `aweme_id`, `desc`, `author`, `music`, `cha_list`, `video`, `statistics`, `status` and `share_url`. Engagement statistics can include likes, comments, shares, plays and collections; video duration is measured in milliseconds. The data schema is flexible and may be null, so inspect the actual response rather than assuming every result has every field.
ae_xiaohongshu_pgy_creator_note_list_pro
获取小红书蒲公英创作者的笔记列表详情,支持按广告类型、笔记形式(图文/视频/合作)以及阅读量或互动量排序进行分页查询。参数请放在必填的 request_body 对象中。is_third_platform:1=仅跨域合作笔记,0=不限。
ae_tiktok_search_users
Search for users by a required keyword. Optionally set offset and count for pagination, filter by follower-count range or verified profile type, and prioritize username relevance. The response contains search data in `data` and metadata such as status code, request parameters, timestamp, and request identifier.