agentleFS
nikhonitnikhonit/social-media-api-skills/skills/tiktok-api/SKILL.md

tiktok-api

Skill3 starsChanged 3 months agoMIT-0View on GitHub

TikTok data toolkit via ScraperSocial — video transcripts, AI summaries, post and channel stats, comments, follower lists, song lookups, keyword and hashtag search.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    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. Very specific:

    Every script also accepts --limit, --cursor, --fields, --format and --fresh.

    line 90
  2. Distinctive:

    Every response uses the same envelope:

    line 98
  3. Very specific:

    Run get_transcript.js with the video URL.

    line 133

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “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 API skill

Read public TikTok data as clean JSON. This TikTok API skill turns a video URL into a full transcript or an AI summary, pulls engagement stats for any post or creator, walks comment threads, lists follower and following graphs, and searches TikTok by keyword, hashtag, user or song — without a headless browser, a login, or a scraping stack to maintain.

The transcript endpoint is the one most people come for: give it a video URL and it returns the spoken words, which is what makes TikTok content searchable, summarisable and usable as model input.

This skill wraps 16 live TikTok endpoints from the ScraperSocial API. Public data only, returned as clean JSON.

When to use this skill

  • A user gives you a TikTok video URL and asks what is said in it, or asks for a summary.
  • You are researching a creator: follower counts, posting cadence, which videos performed.
  • You need the comments on a video, for sentiment or for finding recurring questions.
  • You are tracking a hashtag, a trend or a sound and want the videos ranked under it.
  • You want to find creators or videos by keyword rather than by URL.

Do not use this skill when

  • A TikTok link appears in passing and the user has not asked anything about it. Every call spends credits — a mention is not a request.
  • The user asks something answerable from the page they already pasted. If they pasted the caption and want it rephrased, no API call is needed.
  • You need private data: non-public accounts, DMs, analytics only the account owner sees, or anything behind a login. The API returns public data only and will 404 rather than guess.
  • You need to post, comment, follow or upload. This is read-only.
  • You need a guaranteed-complete follower list for a large account. Follower endpoints are paginated and charged per item; pulling millions of rows is slow and expensive. Sample, or ask the user first.

Setup

export SCRAPERSOCIAL_KEY=sk_live_...   # https://scrapersocial.com/app/keys

Signing up at scrapersocial.com/signup includes 100 free credits, no card required.

Permissions and scope

These scripts do exactly three things and nothing else: read one environment variable (SCRAPERSOCIAL_KEY), send HTTPS requests to https://api.scrapersocial.com (the host is a constant in the code, not configurable), and print JSON. They run as node scripts/<name>.js with no other shell use, no file writes and no persistence. call_endpoint.js accepts only the 16 documented /v1/tiktok/ paths in the table below and refuses anything else before a request is made.

Scripts

ScriptWhat it doesInputCredits
get_channel.js --section=repostsList TikTok channel reposts.handle3 credits per video
get_channel.js --section=statsFollower counts and profile metadata for a TikTok account.handle3 credits
get_channel.js --section=top-videosList TikTok channel top videos.handle3 credits per video
get_channel.js --section=videosLatest videos from a TikTok account with per-video metrics.handle3 credits per video
get_audience.js --section=followersList TikTok followers.handle3 credits per profile
get_audience.js --section=followingList TikTok following.handle3 credits per profile
search.jsKeyword search over TikTok videos.query4 credits per result
search.js --section=hashtag-searchVideos posted under a TikTok hashtag.query3 credits per result
search.js --section=top-searchList TikTok top search.query4 credits per result
search.js --section=user-searchList TikTok user search.query3 credits per profile
get_song.jsGet TikTok song.url3 credits
get_song.js --section=videosList TikTok song videos.url3 credits per video
get_transcript.jsFull transcript of a TikTok video with timestamped segments.url8 credits
get_transcript.js --section=summaryShort AI summary of a TikTok video, built on its transcript.url10 credits
get_comments.jsPaginated comments on a TikTok video; charged per comment returned.url2 credits per comment
get_stats.jsViews, likes, comments and shares for a single TikTok video.url3 credits
call_endpoint.jsCall any /v1/tiktok/ endpoint directlypath + paramsvaries

Every script also accepts --limit, --cursor, --fields, --format and --fresh. --limit is clamped to 50: above that the API returns an async job instead of data.

Example

node skills/tiktok-api/scripts/search.js "sourdough starter"

Every response uses the same envelope:

{ "data": {}, "meta": { "count": 20, "next_cursor": "..." }, "request_id": "a18e4ef88ad4b00b" }

meta appears on paginated endpoints only. Keep request_id — support traces calls with it.

Endpoints

EndpointInputReturnsCreditsPaginated
/v1/tiktok/channel-repostshandlevideo3 credits per videoyes
/v1/tiktok/channel-statshandleprofile3 creditsno
/v1/tiktok/channel-top-videoshandlevideo3 credits per videoyes
/v1/tiktok/channel-videoshandlevideo3 credits per videoyes
/v1/tiktok/commentsurlcomment2 credits per commentyes
/v1/tiktok/followershandleprofile3 credits per profileyes
/v1/tiktok/followinghandleprofile3 credits per profileyes
/v1/tiktok/hashtag-searchqueryresult3 credits per resultyes
/v1/tiktok/searchqueryresult4 credits per resultyes
/v1/tiktok/songurlsong3 creditsno
/v1/tiktok/song-videosurlvideo3 credits per videoyes
/v1/tiktok/statsurlvideo3 creditsno
/v1/tiktok/summaryurlvideo10 creditsno
/v1/tiktok/top-searchqueryresult4 credits per resultyes
/v1/tiktok/transcripturlvideo8 creditsno
/v1/tiktok/user-searchqueryprofile3 credits per profileyes

Costs shown per item for paginated endpoints, per call otherwise. Failed calls are not charged.

FAQ

How do I get a TikTok transcript from a URL?

Run get_transcript.js with the video URL. It returns the spoken content as text, 8 credits per video. Videos with no speech return an empty transcript rather than an error.

What is the difference between transcript and summary?

--section on the same script: the default returns the raw transcript, --section=summary returns an AI-written summary of the video. Summary costs 10 credits and reads the video itself, so you do not need to fetch the transcript first.

Can I search TikTok without a video URL?

Yes. search.js covers keyword search, --section=top-search for top results, --section=hashtag-search for a hashtag, and --section=user-search for creators.

Is this an official TikTok API?

No. ScraperSocial is independent and not affiliated with TikTok. It returns publicly available data only. For posting, ad management or owned-account analytics, use TikTok's own developer platform.

What does it cost?

Between 2 and 10 credits depending on the endpoint — the table above lists each one. Signup includes 100 free credits with no card, which is enough to try every endpoint here.

Links

Disclaimer

ScraperSocial is an independent product and is not affiliated with, endorsed by, or sponsored by any platform named in this repository. All product names, logos and brands are property of their respective owners. These skills return only publicly available data.

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.