TikTok KeyAPI Content Analysis

Problem

When an agent needs to query TikTok creators, products, videos, livestreams, trends, or ad insights, the practical issue is often not making API calls, but mapping a natural-language goal to the right endpoint without guessing paths, parameters, or response shapes from memory. keyapi-tiktok constrains this into a KeyAPI REST workflow: start from the business goal, choose a scenario, verify the current docs, then execute the request instead of scraping pages or assembling routes ad hoc.

How it works

The skill compresses the request into a few decision fields: Goal, Entity, Scope, Metric or sort, Pagination depth, and Output format. It uses references/scenarios.md to select an influencer, Shop, content, or intelligence scenario, then applies references/routing-policy.md to decide whether the task is search, detail, resolver, ranking, or a composed flow. Before any live call, it prefers search-keyapi-docs.mjs or the official docs.keyapi.ai/llms.txt index to verify the current method, /v1/... path, parameters, and response contract; after that, it executes the request with keyapi-api.mjs, or falls back to a host HTTP client using the same documented REST request.

Boundaries

This skill fits structured TikTok data retrieval, comparison, and reporting when KeyAPI credentials are available and the official docs are current. It does not replace manual platform browsing or unauthenticated data access. If skill guidance conflicts with live documentation, the official docs take precedence. For broad reports or multiple adjacent calls, confirm sections, scope, and pagination depth before running a multi-endpoint workflow.