
Figmirror
When you screenshot a figure from a NeurIPS paper and want your own data plotted in that exact visual style, this skill runs a three-agent loop: a preprocessor crops the reference image, a drawer writes matplotlib code matching the style, and a vision-only reviewer audits the output against the reference without seeing your data. It iterates up to six times or until the reviewer says ship. The output is a self-contained Python script with embedded data and a type-42 PDF. It explicitly does not trigger on generic matplotlib requests, only when you provide both a reference image and your own data. The architecture is careful about separating style transfer from layout cloning, so it matches colors and typography without copying axis limits that don't fit your numbers.
- GitHub stars
- 457
- last updated
- May 25, 2026
Install
$ npx -y skills add vila-lab/figmirror --skill figmirror --agent claude-codeInstalls into .claude/skills of the current project.
Files
FigMirror (figmirror)
Use this skill when the user wants to:
- Transfer the visual style of a top-conference paper figure to their own data.
- Produce a camera-ready matplotlib figure matching a reference screenshot in style, not in data.
- Mirror 3D paper-figure references such as surfaces, scatter, trajectories, bars, layered waterfalls, or plane projections when the reference or data is actually 3D.
- Receive a self-contained
.pyscript with editable inline data plus PNG/PDF outputs.
Required Inputs
- A reference figure screenshot (
PNG/JPG). It may include margins, captions, neighboring panels, or page text; Stage 0 preprocesses it. - The user's data in any parseable form: pasted table, CSV, TSV, markdown table, or dirty terminal text.
- A working directory for iteration artifacts.
3D Insert Gate
Enable references/three-d-prompting.md only when the user asks for a 3D
figure, the reference is visibly 3D, or the parsed data requires a 3D encoding
such as x/y/z, surfaces, trajectories, layered profiles, closed objects, 3D
small multiples, 3D bars, or plane projections. Do not use this insert to turn
an ordinary 2D task into 3D.
Sponsors
Details
- Repository
- github.com/vila-lab/figmirror
- Categories
- Code Review & Quality, Data Science & ML, Release Management, Design & UI/UX
- First seen
- Jun 11, 2026