Influencer Fit Analyzer

What problem it solves

When the task is not diagnosing one account, but choosing a short list of creators from a category, budget band, region, or a batch of candidate profiles, the practical question is: who is worth contacting? This skill turns scattered platform profiles, pasted bios, and content evidence into an 8–12 person shortlist memo. It does not write outreach copy or manage posting; it produces a defensible talk or no-talk filter.

How it works

The skill is driven by three core inputs: category, budget or follower band, and region. If the user already has account links, @ handles, bios, or post samples, it can skip public lookup and work directly from that material. The core loop is:
- Collect candidates: from pasted inputs or from confirmed, optional public queries.
- Normalize evidence: capture follower counts, recent median views, engagement, and content pillars, clearly labeling data as looked-up or user-provided.
- Score and narrow: compare category fit, budget fit, market, recent performance, and engagement to reach 8–12 candidates.
- Deliver the memo: include a go/no-go recommendation, and when queries were run, retain the task ID, final state, and billing.net_charged_credits.
Optional public reads may cover TikTok, Douyin, Xiaohongshu, Instagram, YouTube, and X, but every lookup is a separate paid action that must be confirmed before execution. When no lookup is performed, the skill does not estimate missing numbers or invent performance data.

Boundaries and cautions

It is suited for candidate filtering, not full account diagnosis, first-post creation, UGC ad production, or paid social asset workflow. Inferences must be marked as inferences with evidence, and missing metrics are explicitly left missing. X does not support user search, YouTube requires a channel URL, and changing query parameters or paginating creates a new paid request.