Photo to 3D

One-click photo to 3D model pipeline. Upload any photo, AI generates a clean isometric view via Gemini (Nano Banana), then converts it to a production-ready...

Security Scan

The skill's code and instructions are coherent with its stated purpose: it preprocesses an image via Gemini then uploads to Tripo3D to produce a .glb, and the required API keys match those services.

Purpose & Capability

The name/description require Gemini and Tripo3D access; the script uses GEMINI_API_KEY and TRIPO_API_KEY and calls the expected endpoints. Required credentials and network calls are proportionate to converting images → isometric view → 3D model.

Instruction Scope

Instructions and the script operate within the stated pipeline: they read the input image, call Gemini and Tripo3D APIs, and write output files. Important privacy note: the user's image is uploaded to external services (Google generativelanguage and api.tripo3d.ai); this is expected but should be considered before use.

Install Mechanism

No install spec (instruction-only with an included script) — low install risk. One practical omission: the script uses the Python 'requests' library but the skill metadata does not declare dependencies or give an install step; users must ensure Python and requests are available.

Credentials

Only GEMINI_API_KEY and TRIPO_API_KEY are required, which directly match the two external services used. No unrelated secrets or config paths are requested.

Persistence & Privilege

always is false, the skill does not request persistent or elevated agent privileges, and it does not modify other skills or system-wide settings.

Assessment

This skill appears internally consistent and does what it describes, but review these points before installing: - The script uploads your images to two external services (Google generativelanguage and Tripo3D). Do not use sensitive or private images unless you accept those services' privacy/retention policies. - You must provide GEMINI_API_KEY and TRIPO_API_KEY; check billing and permissions for those keys. The script prints small token fragments in logs (not full keys) but network requests will include your keys — treat them as sensitive. - Ensure your environment has Python and the 'requests' library installed (the skill does not include dependency installation instructions). - Origin is unknown and there's no homepage; if you need higher assurance, request provenance or a vetted package/repository for the code, and consider running it in an isolated environment first. - If you plan to run at scale, verify Tripo3D's rate/usage limits and costs and confirm the Gemini usage method matches your Google/API account setup.

Photo to 3D Model

Two-step pipeline that converts any photo into a 3D model:

1. Gemini preprocess — Transform photo into a clean white-background 45° isometric view
2. Tripo3D generate — Convert the isometric image into a .glb 3D model

Requirements

• GEMINI_API_KEY — Get from https://aistudio.google.com/apikey
• TRIPO_API_KEY — Get from https://platform.tripo3d.ai/

Usage

# Full pipeline: photo → isometric view → 3D model
python3 scripts/photo_to_3d.py <image_path>

# Custom prompt for Gemini preprocessing
python3 scripts/photo_to_3d.py <image_path> --prompt "your custom prompt"

# Skip preprocessing (image is already a clean isometric view)
python3 scripts/photo_to_3d.py <image_path> --skip-preprocess

# Custom output directory
python3 scripts/photo_to_3d.py <image_path> --output-dir ./my_output

Output

• output/{name}_isometric.png — Gemini-generated isometric view
• output/{name}_model.glb — Final 3D model file

Default Gemini Prompt

The built-in prompt generates a 45° isometric "3D-printed model" style render with PBR materials on a pure white background. Override with --prompt for custom styles (e.g., game assets, architectural models, product renders).

Notes

• Supported input: .jpg, .jpeg, .png, .webp
• Tripo3D generation takes ~1-3 minutes depending on complexity
• For higher quality, preprocess source images with upscayl before running
• Output .glb files can be converted to .usdz for Apple AR using usdzconvert