WeChat Article Visual Story

The Problem It Solves

Creating a WeChat article often involves disjointed work on copy and visuals, leading to inconsistent styles or mismatches between text and images. Manually adjusting cover ratios and in-article image placements adds overhead before publication.

How the Skill Works

This skill integrates copy generation and image creation into a streamlined pipeline:

• End-to-End Workflow: From a user-provided topic or draft, it generates candidate titles, summaries, and body text, then plans corresponding in-article images for each segment, ensuring visuals directly support content rather than serving as decoration.
• Core Toolchain: Uses beatra.images.generate or beatra.images.transform for rendering, with all remote calls executed via the bundled client scripts/mcp_client.py, avoiding direct REST exposure.
• Confirmation and Control: Paid image generation only starts after copy approval; separate confirmation cards are issued for cover and in-article images, allowing parameter changes (e.g., cover text or composition) but requiring re-approval for each modification.

Applicability and Considerations

• Scope: Designed for fresh creation from topic to final article; for existing copy needing images, use wechat-article-visual-pack; for covers only, use wechat-cover-maker.
• Content Accuracy: Product details, pricing, and claims in the article must come from the user; the skill avoids generating plausible but unverified data to prevent misleading readers.
• Visual Consistency: During delivery, verify that all images share a cohesive visual style and that the cover is readable at feed dimensions—especially when text is rendered, as the output is an AI-generated image with potential rendering errors.