image-to-relief-stl

What This Skill Does

The image-to-relief-stl skill is a specialized utility designed for OpenClaw agents to bridge the gap between 2D generative AI and physical 3D manufacturing. It functions as a deterministic transformation pipeline that takes a flat image file and converts its color or grayscale data into a physical heightmap, ultimately exporting a watertight, 3D-printable STL file. This is ideal for creating custom bas-relief sculptures, physical badges, or textured prototypes directly from digital artwork. By interpreting pixels as Z-axis elevation data, the skill automates the otherwise tedious CAD process of extrusion and surface generation.

Installation

To integrate this tool into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/ajmwagar/image-to-relief-stl. Ensure you have the necessary system dependencies installed, particularly those handling image processing and mesh generation (as noted in the source repository: openclaw/skills). Verify installation by attempting a test run with a simple sample image using the provided shell script wrapper.

Use Cases

This skill is perfect for scenarios involving rapid physical prototyping. For instance, you can use it to create unique tactile labels for accessible hardware, custom coins or medallions generated from iconography, or decorative wall art created from stylized flat-vector illustrations. It is specifically optimized for agents that first generate assets via models like nano-banana-pro, creating a full end-to-end automation loop from prompt to physical file.

Example Prompts

1. "Take the image I just generated of the mountain range, convert it to a bas-relief STL with a base thickness of 2mm, and save it as mountain_relief.stl."
2. "Use the grayscale mode to turn my logo.png into a printable relief. Set the max height to 5mm and make sure the pixel resolution is set to 0.2mm for fine detail."
3. "Convert this color-segmented mascot image into a 3D model using a palette where the background is 0mm and the character is 4mm high. Output the result to model.stl."

Tips & Limitations

For optimal results, ensure your source images are clean. When using palette mode, avoid gradients, shadows, and anti-aliasing; these elements introduce noise into the heightmap that can result in messy, unprintable geometry. For the best physical output, use bold, distinct shapes with sharp edges. Keep in mind that the --pixel parameter is a direct trade-off between detail and file size: smaller values create higher-resolution meshes but exponentially increase processing time and file size, which can affect slicer performance. Finally, verify your source image is watertight—meaning no gaps or floating segments—before conversion to ensure the resulting STL remains a single, coherent mesh.