What problem does it solve?
Turning an HTML slide design into a PowerPoint file usually means pasting screenshots that cannot be edited. This Skill parses structured HTML slides into a semantic model and rebuilds them as native PowerPoint text boxes, shapes, chips, arrows, and panels, so every element stays editable.
Core Features & Use Cases
- Preset-driven conversion: Maps known slide families (v9-architecture, ai-runtime-page) from HTML into native PPTX objects using cheerio for DOM extraction and pptxgenjs for rendering.
- Automatic layout fitting: Computes font sizes and box heights per text block to reduce overflow, then runs a preflight QA script that flags spacing and overflow risks with severity levels.
- Extensible preset system: When HTML does not match an existing slide family, you add a new preset covering DOM extraction, layout mapping, and QA rules instead of forcing arbitrary CSS through.
- Use Case: You have an analyst-style architecture page designed in HTML and need an editable .pptx for a client deck. Run the converter with the matching preset, review the QA report, and iterate until spacing looks right.
Quick Start
Ask the agent to convert your HTML slide file into an editable PPTX using the matching preset, then run the preflight QA check on the dumped model.