Image To Relief Stl

Turn a source image (or multi-color mask image) into a 3D-printable bas-relief STL by mapping colors (or grayscale) to heights. Use when you have an image from an image-gen skill (nano-banana-pro, etc.) and want a real, printable model (STL) via a deterministic pipeline.

Security Scan

The skill's files, required tools, and runtime instructions are consistent with its stated purpose (converting images to bas-relief STL); nothing appears to request unrelated credentials or perform unexpected I/O or network exfiltration.

Purpose & Capability

Name/description match the actual requirements: python3 plus potrace/mkbitmap are reasonable for raster processing and optional SVG preview. The included scripts implement a local heightfield→STL pipeline consistent with the description.

Instruction Scope

SKILL.md and scripts confine actions to converting the provided input image to STL and optionally generating a potrace SVG preview. The scripts read only the supplied input image, write the specified output files, and do not reference unrelated system files or external endpoints.

Install Mechanism

The install entries are apt/brew for potrace (a legitimate source for the required binaries). There are no obscure download URLs or archive extraction steps. The runtime creates a local Python venv and installs Pillow via pip (expected for image handling).

Credentials

No credentials or sensitive environment variables are requested. The scripts use XDG_CACHE_HOME or HOME to place a per-skill venv, which is reasonable and explained in the wrapper script.

Persistence & Privilege

The wrapper creates a persistent venv under ${XDG_CACHE_HOME:-$HOME/.cache}/agent-skills which survives across runs and will pip-install Pillow (network fetch). This is normal for convenience but means the skill will store installed packages on disk.

Assessment

This skill appears to do exactly what it says: convert an image to a bas-relief STL. Before installing, be aware that: (1) it expects potrace/mkbitmap available via apt or brew, (2) the wrapper will create a venv in ~/.cache (or $XDG_CACHE_HOME) and pip-install Pillow (which requires network access to PyPI), and (3) outputs are written to the paths you supply. If you’re comfortable allowing a local venv and a brief pip install, the skill is reasonable to use. If you prefer, you can preinstall Pillow and the potrace package to avoid the wrapper’s pip/network activity.

Generate a watertight, printable STL from an input image by mapping colors (or grayscale) to heights.

This is an orchestrator-friendly workflow:

• Use nano-banana-pro (or any image model) to generate a flat-color image.
• Run this skill to convert it into a bas-relief model.

Practical constraints (to make it work well)

Ask the image model for:

• exactly N solid colors (no gradients)
• no shadows / no antialiasing
• bold shapes with clear edges

That makes segmentation reliable.

Quick start (given an image)

bash scripts/image_to_relief.sh input.png --out out.stl \
  --mode palette \
  --palette '#000000=3.0,#ffffff=0.0' \
  --base 1.5 \
  --pixel 0.4

Grayscale mode

bash scripts/image_to_relief.sh input.png --out out.stl \
  --mode grayscale \
  --min-height 0.0 \
  --max-height 3.0 \
  --base 1.5 \
  --pixel 0.4

Outputs

• out.stl (ASCII STL)
• optional out-preview.svg (vector preview via potrace; best-effort)

Notes

• This v0 uses a raster heightfield meshing approach (robust, no heavy CAD deps).
• The --pixel parameter controls resolution (smaller = higher detail, bigger STL).