Viral TikTok Pattern Miner
Reverse-engineer the week's winning TikTok hooks, sounds, and formats before they saturate
https://apidirect.io/mcp?token=YOUR_API_KEY
Decode the winning TikTok formats and sounds for {niche} in {region} over the last week so I can brief my editor.
viral-tiktok-pattern-miner.
Any agent can also call get_skill(skill_id="viral-tiktok-pattern-miner") to pull these steps on demand.
What it does
Sorting TikTok by most_liked over a 7-day window surfaces proven winners; cross-checking the most_recent feed reveals which of those formats and sounds are still fresh enough to ride rather than already played out.
Who it's for: Short-form content strategists and TikTok editors
Inputs
| Input | Required | Description | Example |
|---|---|---|---|
niche |
Yes | The content niche to mine for winning patterns | home espresso |
region |
No | Two-letter region code for the TikTok feed | us |
How your agent runs it
-
1
search_tiktok(query={niche}, sort_by=most_liked, publish_time=7, region={region}, get_sentiment=true)Take the top-liked videos of the week and log each one's music_title/author, hook, and format, keeping positively-received winners.
-
2
search_tiktok(query={niche}, sort_by=most_recent, publish_time=7, region={region})Compare against the freshest uploads to flag which winning formats are still emerging versus already saturated.
-
3
search_tiktok_users(query=<winning_creator>, pages=2)Resolve the creators behind the top videos and capture follower and video counts to identify who to model or partner with.
Tools used
The full playbook
This is exactly what the MCP returns to your agent (via the viral-tiktok-pattern-miner prompt or get_skill tool), with your inputs filled in.
SKILL: Viral TikTok Pattern Miner
Sorting TikTok by most_liked over a 7-day window surfaces proven winners; cross-checking the most_recent feed reveals which of those formats and sounds are still fresh enough to ride rather than already played out.
You are running this skill on API Direct via its MCP tools. Execute the steps below yourself by calling the named tools in order — values in <angle brackets> come from a previous step. Then deliver the result described at the end.
INPUTS:
- niche (required): <niche — ASK THE USER>
- region (optional): (optional — e.g. us)
STEPS:
1. Tool `search_tiktok` — search_tiktok(query=<niche>, sort_by=most_liked, publish_time=7, region=<region>, get_sentiment=true)
Take the top-liked videos of the week and log each one's music_title/author, hook, and format, keeping positively-received winners.
2. Tool `search_tiktok` — search_tiktok(query=<niche>, sort_by=most_recent, publish_time=7, region=<region>)
Compare against the freshest uploads to flag which winning formats are still emerging versus already saturated.
3. Tool `search_tiktok_users` — search_tiktok_users(query=<winning_creator>, pages=2)
Resolve the creators behind the top videos and capture follower and video counts to identify who to model or partner with.
DELIVER: A content playbook of the week's winning hooks, sounds, and formats plus the creators driving them.
Note: each underlying tool call is billed at its normal endpoint price; get_sentiment adds a small per-page surcharge. Page through results as needed but stop once you have enough to deliver the outcome.