paper-illustration-image2

To generate academic paper figures, you provide a research figure request that the Skill converts into a staged, style-verified image prompt. It then renders publication-quality raster figures with labeled modules and directed arrows using native codex-image2 bridge rendering.

Yes, you can create CVPR or NeurIPS style architecture diagrams. The Skill applies a strict academic review gate that scores figures against grayscale, print suitability, and conference style constraints to ensure the output meets publication standards.

Yes, image generation includes multi-step layout and style verification. The Skill iteratively improves the module-aware figure prompt and enforces a strict academic review gate with a target score of at least 9 out of 10 before finalizing the output.

You get a LaTeX include snippet by asking the Skill to render a specific workflow diagram. Upon passing PNG verification, it produces a finalized PNG artifact alongside a LaTeX-ready include file and verification logs in the output directory.

The best way to ensure arrow direction and readability is to use the Skill's strict academic review gate. It specifically evaluates arrow direction, thickness, and label readability before emitting the finalized paper artifact to the designated directory.

Yes, you need a local Codex app-server MCP bridge to render native imageGeneration output. The Skill requires this bridge to produce genuine raster figures without any fallback masquerading during the image generation process.