# Model Generation Tools Reference
Quick reference for all model generation MCP tools.
## Overview
The generation tools allow Claude to create 3D models from:
1. **Natural language descriptions** → Parametric generation (CadQuery/OpenSCAD)
2. **Reference images** → Dimension extraction + parametric generation
3. **Text prompts** → AI generation (Meshy, Tripo3D)
4. **Templates** → Pre-built customizable designs
## Tools
### `bambustudio_parse_requirements`
Parse natural language into structured model parameters.
```json
{
"description": "I need a tube squeezer for a 65mm lotion bottle, heavy duty",
"target_dimension_mm": 65
}
```
**Returns:**
- Category (tube_squeezer, holder, bracket, clip, etc.)
- Extracted dimensions
- Fit type (tight, snug, sliding, loose)
- Strength requirements
- Wall thickness recommendations
---
### `bambustudio_analyze_reference_image`
Extract dimensions from a photo.
```json
{
"image_path": "/path/to/photo_with_ruler.jpg",
"known_dimension_mm": 65 // Optional calibration
}
```
**Features:**
- Automatic ruler/scale detection
- Object edge measurement
- Shape feature detection (slots, holes)
- Category suggestion
---
### `bambustudio_list_templates`
List available model templates.
**Built-in templates:**
| Template | Description | Key Parameters |
|----------|-------------|----------------|
| `tube_squeezer` | Tube/bottle squeezer | tube_diameter, wall_thickness, clearance |
| `phone_holder` | Phone/tablet stand | width, depth, angle |
| `cable_catch` | Cable management clip | cable_diameter, num_slots |
---
### `bambustudio_generate_from_template`
Generate a model from a customizable template.
```json
{
"template_name": "tube_squeezer",
"tube_diameter": 65,
"wall_thickness": 3.0,
"clearance": 1.5
}
```
**Best for:** Functional parts with known dimensions.
---
### `bambustudio_generate_parametric`
Generate from natural language using parametric CAD.
```json
{
"description": "A heavy duty tube squeezer for 65mm lotion bottle with grip texture",
"target_dimension_mm": 65
}
```
**Process:**
1. Parse requirements from description
2. Classify object type
3. Generate CadQuery code (or OpenSCAD fallback)
4. Export to STL
**Best for:** Functional parts described in natural language.
---
### `bambustudio_ai_generate`
Generate using AI text-to-3D services.
```json
{
"prompt": "A cute cartoon cat figurine sitting",
"style": "cartoon"
}
```
**Styles:** realistic, cartoon, sculpture
**Requirements:** API key (MESHY_API_KEY or TRIPO3D_API_KEY)
**Best for:** Artistic/organic shapes, decorative items.
---
### `bambustudio_ai_status`
Check AI generation job status.
```json
{
"job_id": "abc123"
}
```
**Returns:** status (processing/completed/failed), progress %, download URL
---
## Choosing the Right Generation Method
| Use Case | Recommended Tool |
|----------|------------------|
| Tube squeezer with specific dimensions | `generate_from_template` |
| Functional part from description | `generate_parametric` |
| "Make me a phone stand" | `parse_requirements` → `generate_from_template` |
| Artistic figurine | `ai_generate` |
| Dimension from photo | `analyze_reference_image` → `generate_from_template` |
---
## Example Workflow: Lotion Bottle Squeezer
```python
# 1. Parse the request
requirements = await bambustudio_parse_requirements({
"description": "Heavy duty squeezer for Gold Bond lotion, 65mm bottle",
"target_dimension_mm": 65
})
# 2. Generate from template (recommended for precise functional parts)
model = await bambustudio_generate_from_template({
"template_name": "tube_squeezer",
"tube_diameter": 65,
"wall_thickness": 3.0, # Heavy duty
"clearance": 1.5,
})
# 3. Slice
sliced = await bambustudio_slice_model({
"file_path": model["output_path"],
"preset": "tube_squeezer_strong"
})
# 4. Print and iterate...
```
---
## Installing CAD Dependencies
For parametric generation, install CadQuery:
```bash
pip install cadquery-ocp --break-system-packages
```
Or use OpenSCAD as fallback:
```bash
brew install openscad # macOS
```
For AI generation, use 1Password CLI to securely inject your API keys:
```bash
# Store your API key in 1Password, then run with:
op run --env-file=.env.template -- python -m bambustudio_mcp
# Or for one-off commands:
MESHY_API_KEY=$(op read "op://Personal/Meshy/api_key") python your_script.py
```
See the main README for full 1Password CLI setup instructions.
