ads-tiktok is the TikTok worker of the Claude Ads toolkit. It collects the objective, conversion definition, account and campaign age, geography, date window, timezone, currency, spend and targets, then reads the TikTok audit reference and the relevant shared measurement, benchmark, creative, automation, policy and scoring references.
It evaluates the applicable controls, separates observations, diagnoses, recommendations, opportunities and proposed changes, marks uncertainty and contradictions, and returns schema-valid findings to the conductor. Server-side events go to ads-server-side-tracking and campaign changes to ads-optimize.
What You Get
- Findings on Pixel measurement and attribution
- A review of mobile-native creative and audiences
- Smart+, Shop, bidding, budget and pacing observations
- Policy findings
- Observations, diagnoses, recommendations, unscored opportunities, contradictions and missing inputs in the common JSON contracts
Common Use Cases
- Auditing a TikTok account before a budget increase
- Reviewing a TikTok Shop or GMV Max setup
- Checking that the Pixel and attribution settings match the conversion goal
- Assessing mobile-native creative and Spark Ads usage
What Makes It Unique
It is the TikTok-specific worker in a twelve-platform suite and uses the same contracts as the rest. A benchmark is applied only after checking objective, geography, methodology, sample size, conversion lag and account maturity, and the skill refuses to issue universal pause, bid, budget, learning-phase or attribution rules.
Under The Hood
The skill is written instructions with no scripts of its own. The agent reads the main ads contract and the TikTok audit reference in the plugin or standalone install, then works from the exports, screenshots or authenticated reads you give it. It does not write a shared result file or change the account. External content is treated as data, and every account change stays a draft until the main mutation gate passes.