TikTok Account Audit

Audit a TikTok account’s recent posts like a growth analyst and turn the patterns into tests for the next posts.

v0.1.1

View example draft · How to use

Example draft

@duolingo, 20 recent posts

Posts from 2026-10-13 to 11-04 on @duolingo (18.4 M followers).

What works

Trending or other creator’s sound: n 14, median 1,425,250 plays Own original sound: n 5, median 418,197 plays ≤10 s: n 12, median 1,244,895 · >10 s: n 7, median 644,672 The three biggest posts are 5–9 s.

Test to run

Trend-first ratio: post 4 trending-sound posts for every own-sound post for two weeks. Success: the window median above 1.0 M plays.

Excerpt from an actual run on 2026-11-05.

What you get

  • Median plays, engagement and share rate with top and bottom posts
  • Splits by sound, length and format, each with its sample size
  • Hourly views, comment keywords and a fake-view verdict read in context

Good fit

  • Planning the next posts from evidence
  • Checking a creator’s views before a partnership

Not for

  • Watch time, completion or demographics, which only the owner can see
  • Calling an account fraudulent from one score

How to use

An Agent that can read the complete Skill folder. Live research also needs a SandBase connection; supplied material can be used without a lookup.

Download and unpack the complete Skill folder, keep its references, and use it with a local Agent that supports Skills.

Download Skill v0.1.1

Ask your Agent

Use tiktok-account-audit on @duolingo. Analyse the last 20 posts, explain what separates hits from misses, check the top post’s views, and suggest tests for the next two weeks.
  1. Collect posts

    Fetch the profile and at least 20 recent posts; leave pinned posts out of medians.

  2. Find the patterns

    Split by sound, length and format, and compare top and bottom posts.

  3. Deep dive and test

    Analyse the top post, read milestones, and write tests with success measures.

FAQ

How do I use this Skill?

Download and unpack the whole Skill folder, including references, into your Agent’s supported Skill directory. You can also ask the Agent to read its SKILL.md and follow the instructions. Run it locally with your own material.

Does a high fake-view score prove fraud?

No. It is a statistical estimate; fast organic growth also trips its rules, so read it against the account’s history.