To find nice images for the article, I searched around yesterday and finally selected 4 open-source skills from GitHub, and the result feels quite good.

The 4 skills can correspond to different scenarios.

Let me explain, they are not standalone skill projects, but rather within the skills of baoyu and huashu.

Another tip is that using codex to generate images is less effective than invoking gpt-image2; I recommend using gpt-image2.

1. baoyu-image-cards

- Skill Name: `baoyu-image-cards`

- Suitable Scenarios: Social media image cards, knowledge cards, carousel images, tweet illustrations.

- Usage Suggestion: Suitable for breaking down a tweet into 1-3 key cards, highlighting the title, judgement points, and conclusions.

2. baoyu-cover-image

- Skill Name: `baoyu-cover-image`

- Suitable Scenarios: Article cover images, main visuals for tweets, large title images.

- Usage Suggestion: Suitable for use when only one cover image is paired with each tweet, focusing on enlarging the title and core concepts.

3. huashu-wechat-image

- Skill Name: `huashu-wechat-image`

- Suitable Scenarios: Public account illustrations, article illustrations, 16:9 Chinese infographics, explanatory images for the main text.

- Usage Suggestion: Suitable for creating infographics that are accurate in Chinese and have a clear structure, such as comparison charts, flowcharts, and field relationship diagrams.

4. huashu-xhs-image

- Skill Name: `huashu-xhs-image`

- Suitable Scenarios: Xiaohongshu style illustrations, vertical social media cards, large text images for mobile feeds.

- Usage Suggestion: Suitable for creating vertical images, with large titles and fewer highlights, appropriate for mobile reading.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。