agentleFS
aya-talebaya-taleb/video-boost-skills/skills/tiktok-research/SKILL.md

tiktok-research

Skill1 starsChanged 22 days agoMITView on GitHub

Researches what is working on TikTok right now for one topic, market and language - top recent videos and their patterns, current trends, common mistakes, and TikTok's own rules and recommendations - using only sources from the last 12 months (6 with recent), each with a link and date. Use when the user asks what works on TikTok for a topic, why TikTok videos fail, TikTok trends for X, or as part of viral-checklist.

5 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 6 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 4 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write today's date at the top of your notes.

    line 19
  2. Distinctive:

    Check the date on every page.

    line 32
  3. Distinctive:

    Failures (mistakes): Search for why videos in this topic flop, TikTok shadowban myths vs facts, and content TikTok does not recommend.

    line 36
  4. Distinctive:

    Every finding needs a link and a publish date inside the window.

    line 55
  5. Very specific:

    No numbers from memory. If you know something but cannot find a current source, leave it out or put it in gaps.

    line 57
  6. Distinctive:

    Keep claims short and specific.

    line 59
  7. Distinctive:

    Write the notes in plain words.

    line 60
  8. Very specific:

    When done, run source-check on the notes, unless the user asked you not to.

    line 62

“Common” is measured: kept by at least 6 in 10 of the 4 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

TikTok research

Finds what is working on TikTok for one topic, and what is not, with a source and date for every finding.

Inputs

InputDefault
topicask if missing
market (country)US
languageen
window12 months (recent = 6 months)

Write today's date at the top of your notes. Every date check uses it.

Output

notes-tiktok.json in the format in shared/checklist-format.md. Aim for 15 to 30 findings. Fewer is fine if that is all the good evidence there is. List what you could not find under gaps.

Where to look, in this order

  1. TikTok's own rules and tips (official). These beat everything else.
    • Community Guidelines, including the For You feed eligibility standards (what TikTok will not recommend): tiktok.com/community-guidelines
    • How TikTok recommends videos: support.tiktok.com (search "how TikTok recommends content")
    • Creator help and Creator Academy pages, the TikTok Newsroom (newsroom.tiktok.com), and TikTok's yearly trend reports ("What's Next")
    • AI content labels and rules (search "TikTok AI-generated content label") Check the date on every page. Rules change often.
  2. What is trending now (official data). TikTok Creative Center (ads.tiktok.com/business/creativecenter) shows trending hashtags, songs, creators and videos by country and time range. Set the country to the market. If the page will not load in your tools, say so in gaps and do not guess what it shows.
  3. Recent top videos in the topic (examples). Search the web for recent viral or top videos in the topic and market, with the year in the query (for example "travel tips TikTok viral 2026"). Keep only videos where you can see the post date and a view count from the page or a dated article. Note for each: the hook (first 1 to 3 seconds), length, format (talking head, voice over, text on screen, skit, list), on-screen text, sound (trending sound, original voice, music), and call to action.
  4. Studies and analysis (data, analysis). Recent studies with a sample size (for example from social media management or analytics companies), and articles from reputable trade press. Prefer studies that say how many videos or accounts they looked at, which country, and which months.
  5. Failures (mistakes). Search for why videos in this topic flop, TikTok shadowban myths vs facts, and content TikTok does not recommend. A mistake needs a reason and a source, not just an opinion.

What to pull out

Cover each of these, or list it as a gap:

  • Hook: what happens in the first 1 to 3 seconds of top videos
  • Length that works for this topic now (with the source's date)
  • Structure and pacing: cuts, list format, story, payoff at the end
  • On-screen text and captions
  • Sound: original audio vs trending sounds, voice over, music
  • Visual style, faces vs no faces, real footage vs animation
  • Calls to action that work (follow, comment, save, share) and ones that hurt
  • Posting: frequency, times (only with a dated US source), hashtags (how many, which kind)
  • What TikTok will not recommend (For You feed rules), AI labels, copyright and unoriginal content
  • Topic-specific risks (for travel: safety claims, prices that change; for health: no advice)

Rules

  • Every finding needs a link and a publish date inside the window. No date, no finding.
  • Market and language must match, or the finding is marked "scope": "global" (only for platform-wide announcements).
  • No numbers from memory. If you know something but cannot find a current source, leave it out or put it in gaps.
  • Ads results (TikTok's ad studies) are about paid ads. Mark them clearly. Do not present them as organic.
  • Keep claims short and specific. "Hooks matter" is useless; "Top 10 travel tip videos this summer all show the destination in frame 1" is useful.
  • Write the notes in plain words.

When done, run source-check on the notes, unless the user asked you not to.

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.