content-performance-analysis — Claude Code skill: Xiaohongshu post
A skill that breaks down Xiaohongshu post stats: engagement, saves and viral score, plus which format actually performs.
npx skills add vivy-yi/xiaohongshu-skills --skill content-performance-analysis -a claude-codeWhat content-performance-analysis does
The skill kicks in when you need to know why one post pulled views and another flopped. It pulls metrics from the Xiaohongshu Creator Center or Qiangua Data: views, likes, comments, shares, saves, new followers. Then it computes four numbers — engagement rate, save rate, comment rate and viral score (views divided by follower count).
From there it walks through the data step by step: collect metrics, calculate the indicators, sort posts and surface the top and bottom five, extract shared traits from the winners, diagnose the losers. It looks at format (single image, carousel, video), structure, titles, covers, topic and posting time. You get a matrix by content type and a fix list for weak posts.
There is a floor on the data. You need at least 5 published posts, otherwise there is nothing to compare. The skill skips the first hours after publishing and advises waiting 3–7 days. Paid ads are out of scope — use ad analytics for those. Worth checking whether your account has enough history before running this.
Install runs through npx skills add, or gh skill install for GitHub Copilot. The skill file lands in .claude/skills/content-performance-analysis/SKILL.md or the equivalent folder for your platform.
What it helps with
- calculate engagement and save rate for each post
- find shared traits across the top five posts
- work out why some posts underperformed the existing audience
- compare carousels, video and single images by response
- build a fix list for weak posts
Example prompts for the agent
- Analyze my recent Xiaohongshu posts and tell me which format performs best
- Calculate the viral score for every post and flag the underperformers
- Compare carousels and video by engagement over the last 30 days
How to install content-performance-analysis
1Skills CLI (Vercel)
npx skills add vivy-yi/xiaohongshu-skills --skill content-performance-analysis -a claude-code2Manually: put the files in place
~/.claude/skills/content-performance-analysis/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/content-performance-analysis/
1Skills CLI (Vercel)
npx skills add vivy-yi/xiaohongshu-skills --skill content-performance-analysis -a codex2Manually: put the files in place
.agents/skills/content-performance-analysis/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add vivy-yi/xiaohongshu-skills --skill content-performance-analysis -a cursor2Manually: put the files in place
.cursor/skills/content-performance-analysis/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install vivy-yi/xiaohongshu-skills content-performance-analysis2Manually: put the files in place
.github/skills/content-performance-analysis/SKILL.md1Skills CLI (Vercel)
npx skills add vivy-yi/xiaohongshu-skills --skill content-performance-analysis -a gemini-cli2Platform CLI
gemini skills install https://github.com/vivy-yi/xiaohongshu-skills --path skills/03-数据分析/content-performance-analysisManually: put the files in place
.devin/skills/content-performance-analysis/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add vivy-yi/xiaohongshu-skills --skill content-performance-analysis -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about content-performance-analysis
How do I install content-performance-analysis in Claude?
Run npx skills add vivy-yi/xiaohongshu-skills --skill content-performance-analysis -a claude-code. The skill lands in ~/.claude/skills/content-performance-analysis/SKILL.md.
What does content-performance-analysis do?
It computes Xiaohongshu post metrics — engagement, saves, comments, viral score — and finds patterns across strong and weak posts.
When does the skill not apply?
With fewer than 5 published posts, and during the first hours after publishing: the skill advises waiting 3–7 days. It is not built for paid ad analysis.