Generate Xiaohongshu (小红书) blog content and infographic images from a GitHub PR. Use when user provides a GitHub PR URL or asks to create XHS content from a PR.
Generate Xiaohongshu blog content and infographic images from a GitHub PR.
/pr-to-xhs <github-pr-url>
/pr-to-xhs https://github.com/vllm-project/vllm-omni/pull/1689
gh pr view <pr-number> --repo <owner/repo> --json title,body,author,additions,deletions,files,commits
blogs/pr<pr-number>/xiaohongshu/images/xhs-images/<topic-slug>/
├── source-<slug>.md # Source content
├── analysis.md # Content analysis
├── outline.md # Final outline
├── post.json # XHS post configuration
├── prompts/ # Image prompts
│ ├── 01-cover-<slug>.md
│ ├── 02-content-<slug>.md
│ └── ...
├── 01-cover-<slug>.png # Generated images
├── 02-content-<slug>.png
└── ...
Create analysis.md with:
Create outline.md with:
Use glm-image skill (Zhipu AI) to generate:
{
"title": "<title>",
"content": "<body with hashtags>",
"images": ["01-cover.png", "02-content.png", ...],
"tags": ["tag1", "tag2", ...],
"source": {
"type": "github_pr",
"url": "<pr-url>",
"pr_number": <number>
}
}
Default configuration:
Input:
/pr-to-xhs https://github.com/vllm-project/vllm-omni/pull/1689
Output:
blogs/pr1689/xiaohongshu/images/xhs-images/hunyuan-image3-npu-support/