Kuaishou Short Video Operations Growth Assistant

Problem

After a Kuaishou short video is published, operators often review comments in fragments and struggle to answer concrete questions: Is the overall reception positive, contested, or risky? What are users actually asking about? Which comments contain conversion intent or follow-up topic clues? Manually scanning comments is slow and can mistake isolated reactions for the broader trend. kuaishou-sentiment-dashboard turns “Kuaishou short-video comment analysis” into a repeatable task input.

How It Works

The core input is videoUrl, and it supports pasting a Kuaishou share link, video page link, or video ID directly. When scripts/run.py is called, the script automatically completes three steps: parsing the share link, creating the analysis task, and polling until the task finishes. The output is oriented toward operational review rather than a raw data report: it first provides an overall comment-section conclusion to help judge content reputation, discussion focus, and potential public-opinion risk; then it breaks down high-frequency requests, representative comments, sentiment tendencies, and user persona clues; finally it lands on operational suggestions, such as the next content topic, comment-reply strategy, conversion follow-up, or series direction.

Boundaries

This skill is suitable for creators, operators, brand teams, or e-commerce teams that already have a specific Kuaishou content object and need to review comment feedback quickly. It depends on an external service call, so authentication information such as AISKILLS_API_KEY must be prepared before running it. The provided materials do not describe local offline analysis, batch mining of historical videos, or private message/order pipeline analysis, so it should not be treated as a system that can directly read all private Kuaishou backend data.