GEO Optimization with Self-Growth

Problem

Traditional SEO often focuses on keyword coverage and tags, but AI search depends more on question coverage, factual density, and clear citability. When content is marketing-heavy, lacks structured facts, or stuffs keywords, AI engines such as Doubao, Baidu AI, Kimi, and Yuanbao may have difficulty citing it directly. This skill shifts the goal from being found in platform search to being cited by AI answers, while still supporting internal search on Xiaohongshu, Douyin, Official Accounts, and similar platforms.

How it works

The workflow first reads geo-memory/wisdom.md, then analyzes platform, content type, industry, user goal, and target AI platforms. GEO diagnosis dynamically checks title question coverage, factual density, follow-up question coverage, clarity, and retrievability; it also marks keyword stuffing as PASS/FAIL. If the publishing platform supports internal search, it adds optional SEO checks for keyword breadth, title keyword position, tags, and search-suggestion alignment. However, GEO takes priority when the two conflict. The skill then rewrites the title, body, and tags, adding structured data, definitive language, and inline FAQs, with a concise reason for each change. After execution, it reflects on the optimization and only proposes a wisdom.md update when a new pattern, variant, contradiction, or cross-domain transferable insight appears.

Boundaries

It is suitable for service promotion, product recommendations, knowledge content, guides, and travel notes that can be cited in AI Q&A. It does not promise specific uplift numbers or claim an unverified improvement percentage. It does not modify files without confirmation. If the input contains factual errors, it flags them before optimizing. When a platform SEO tactic conflicts with GEO evidence, especially keyword stuffing, GEO takes precedence.