WeChat Official Account Similar Account Recommender

Problem

When creators choose positioning, topics, or placements, they often lack comparable accounts in the same niche. Searching by name yields scattered signals, while natural-language category queries can miss the platform’s internal taxonomy. This skill turns “similar account discovery” into a structured query: provide an account name, ID, or category and receive comparable basic information, the latest five articles, peer benchmarks, and top-tier benchmarks.

How It Works

Core capabilities include account diagnostics, peer benchmarking, top benchmarking, and recommendation reasoning. The skill calls the RedFox Data API by name, ID, or category and returns currentAccount, benchmarkAccounts, and topAccounts. Key steps are:
- Show the queried account’s basic information, RedFox Index, and recent five articles;
- Filter peer accounts using reads, interaction rate, share rate, and publishing cadence;
- Recommend mature accounts with roughly 3–5x the reads as catch-up benchmarks;
- Generate grounded reasons using the three-layer weighted model.

Boundaries And Notes

The skill depends on RedFox Data API account coverage and credit balance, so new accounts, low-volume accounts, or sparse-data accounts may return incomplete results. --account_type can be unstable for natural-language categories, so it is better to query a representative account name in that domain and then confirm the returned account category. Interaction rates above 100% are treated as anomalies and are not displayed directly in the recommendation.