product-image-processor

Name: product-image-processor
Author: AlpacaLabsLLC

260stars54forks

How to install product-image-processor?

alpacalabsllc/skills-for-architects/product-image-processor

$npx -y skills add AlpacaLabsLLC/skills-for-architects --skill product-image-processor

Installs into the current project.

Summary

Download, resize, and remove backgrounds from product images at scale. Use when the user asks to "process product images", batch-download images from the schedule, strip backgrounds, or standardize product photos.

Files · 1

View on GitHub

SKILL.md

1	# /product-image-processor — Product Image Processor
2
3	Download product images from a Google Sheet, normalize sizing, and remove backgrounds. Saves output at each processing stage.
4
5	Works with the master Google Sheet — the 33-column schema defined in `../../schema/product-schema.md`. Image URLs are in column AC, product names in column E. Read `../../schema/sheet-conventions.md` for CRUD patterns with MCP tools.
6
7	## Step 1: Get Input
8
9	If no arguments provided, ask the user:
10	1. Spreadsheet ID — the Google Sheets ID (from the URL: `docs.google.com/spreadsheets/d/{ID}/...`). 2. Image URL column — which column contains image URLs (default: `AC` in the master schema, or the user can specify)
11	3. Name column (optional) — which column has product names for file naming (default: `E` in the master schema). If not provided, derive names from the image URL/filename.
12	4. Output location — where to save the images. Suggest `./product-images-YYYY-MM-DD/` as default but let the user pick any path.
13	5. Header row — whether row 1 is a header (default: yes, row 2 in master schema)
14
15	## Step 2: Read URLs from Google Sheet
16
17	Use `mcp__google-sheets__list_sheets` to inspect the sheet, then `mcp__google-sheets__get_sheet_data` to read the image URL column and optional name column.
18
19	Build a list of `{ index, url, name }` entries. Skip empty rows.
20
21	## Step 3: Create Output Folders
22
23	Create the output directory at the user's chosen path with 3 subfolders:
24
25	```
26	<output-path>/
27	├── originals/ # Raw downloads
28	├── resized/ # Normalized sizing
29	└── nobg/ # Background removed
30	```
31
32	If the folder already exists, append a suffix: `-2`, `-3`, etc.
33
34	## Step 4: Download Images
35
36	Download each image using `curl` in Bash:
37
38	```bash
39	curl -L -o "<output-path>" "<url>"
40	```
41
42	IMPORTANT: Use `curl`, NOT WebFetch. WebFetch processes content through an AI model which corrupts binary image data.
43
44	Name files as: `001-product-name.png`, `002-product-name.png`, etc.
45	- Slugify the product name: lowercase, replace spaces/special chars with hyphens, strip consecutive hyphens
46	- If no name column, extract a name from the URL filename (strip extension and query params)
47	- If the URL gives no usable name, use `001-image.png`, `002-image.png`, etc.
48
49	If the downloaded file is not a PNG (check extension or content type), convert it to PNG during the resize step.
50
51	## Step 5: Resize Images
52
53	Run a Python script to resize all images in `originals/` → `resized/`:
54
55	```python
56	from PIL import Image
57	import os, sys
58
59	input_dir = sys.argv[1] # originals/
60	output_dir = sys.argv[2] # resized/
61	max_edge = int(sys.argv[3]) if len(sys.argv) > 3 else 2000
62
63	for fname in sorted(os.listdir(input_dir)):
64	if not fname.lower().endswith(('.png', '.jpg', '.jpeg', '.webp', '.gif', '.bmp', '.tiff')):
65	continue
66	try:
67	img = Image.open(os.path.join(input_dir, fname))
68	img = img.convert("RGBA")
69	w, h = img.size
70	longest = max(w, h)
71	if longest > max_edge:
72	scale = max_edge / longest
73	new_w, new_h = int(w * scale), int(h * scale)
74	img = img.resize((new_w, new_h), Image.LANCZOS)
75	out_name = os.path.splitext(fname)[0] + ".png"
76	img.save(os.path.join(output_dir, out_name), "PNG")
77	print(f"OK: {fname} → {out_name} ({img.size[0]}x{img.size[1]})")
78	except Exception as e:
79	print(f"FAIL: {fname} — {e}")
80	```
81
82	Rules:
83	- Max 2000px on the longest edge (configurable if user requests)
84	- Preserve aspect ratio
85	- Do NOT upscale — if already smaller than max, keep original dimensions
86	- Convert everything to PNG (RGBA mode for transparency support)
87
88	## Step 6: Remove Backgrounds
89
90	Check if `rembg` is installed. If not, install it:
91
92	```bash
93	pip3 install rembg onnxruntime
94	```
95
96	Then run background removal on all resized images → `nobg/`:
97
98	```python
99	from rembg import remove
100	from PIL import Image
101	import os, sys, io
102
103	input_dir = sys.argv[1] # resized/
104	output_dir = sys.argv[2] # nobg/
105
106	for fname in sorted(os.listdir(input_dir)):
107	if not fname.lower().endswith('.png'):
108	continue
109	try:
110	input_path = os.path.join(input_dir, fname)
111	with open(input_path, 'rb') as f:
112	input_data = f.read()
113	output_data = remove(input_data)
114	img = Image.open(io.BytesIO(output_data))
115	img.save(os.path.join(output_dir, fname), "PNG")
116	print(f"OK: {fname}")
117	except Exception as e:
118	print(f"FAIL: {fname} — {e}")
119	```
120
121	Note: The first run of rembg downloads the u2net model (~170MB). Warn the user

People who liked this also viewed

6 picks

Type

product-image-processor

260stars54forks

View on GitHub

How to install product-image-processor?

alpacalabsllc/skills-for-architects/product-image-processor

$npx -y skills add AlpacaLabsLLC/skills-for-architects --skill product-image-processor

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# /product-image-processor — Product Image Processor
2
3	Download product images from a Google Sheet, normalize sizing, and remove backgrounds. Saves output at each processing stage.
4
5	Works with the master Google Sheet — the 33-column schema defined in `../../schema/product-schema.md`. Image URLs are in column AC, product names in column E. Read `../../schema/sheet-conventions.md` for CRUD patterns with MCP tools.
6
7	## Step 1: Get Input
8
9	If no arguments provided, ask the user:
10	1. Spreadsheet ID — the Google Sheets ID (from the URL: `docs.google.com/spreadsheets/d/{ID}/...`). 2. Image URL column — which column contains image URLs (default: `AC` in the master schema, or the user can specify)
11	3. Name column (optional) — which column has product names for file naming (default: `E` in the master schema). If not provided, derive names from the image URL/filename.
12	4. Output location — where to save the images. Suggest `./product-images-YYYY-MM-DD/` as default but let the user pick any path.
13	5. Header row — whether row 1 is a header (default: yes, row 2 in master schema)
14
15	## Step 2: Read URLs from Google Sheet
16
17	Use `mcp__google-sheets__list_sheets` to inspect the sheet, then `mcp__google-sheets__get_sheet_data` to read the image URL column and optional name column.
18
19	Build a list of `{ index, url, name }` entries. Skip empty rows.
20
21	## Step 3: Create Output Folders
22
23	Create the output directory at the user's chosen path with 3 subfolders:
24
25	```
26	<output-path>/
27	├── originals/ # Raw downloads
28	├── resized/ # Normalized sizing
29	└── nobg/ # Background removed
30	```
31
32	If the folder already exists, append a suffix: `-2`, `-3`, etc.
33
34	## Step 4: Download Images
35
36	Download each image using `curl` in Bash:
37
38	```bash
39	curl -L -o "<output-path>" "<url>"
40	```
41
42	IMPORTANT: Use `curl`, NOT WebFetch. WebFetch processes content through an AI model which corrupts binary image data.
43
44	Name files as: `001-product-name.png`, `002-product-name.png`, etc.
45	- Slugify the product name: lowercase, replace spaces/special chars with hyphens, strip consecutive hyphens
46	- If no name column, extract a name from the URL filename (strip extension and query params)
47	- If the URL gives no usable name, use `001-image.png`, `002-image.png`, etc.
48
49	If the downloaded file is not a PNG (check extension or content type), convert it to PNG during the resize step.
50
51	## Step 5: Resize Images
52
53	Run a Python script to resize all images in `originals/` → `resized/`:
54
55	```python
56	from PIL import Image
57	import os, sys
58
59	input_dir = sys.argv[1] # originals/
60	output_dir = sys.argv[2] # resized/
61	max_edge = int(sys.argv[3]) if len(sys.argv) > 3 else 2000
62
63	for fname in sorted(os.listdir(input_dir)):
64	if not fname.lower().endswith(('.png', '.jpg', '.jpeg', '.webp', '.gif', '.bmp', '.tiff')):
65	continue
66	try:
67	img = Image.open(os.path.join(input_dir, fname))
68	img = img.convert("RGBA")
69	w, h = img.size
70	longest = max(w, h)
71	if longest > max_edge:
72	scale = max_edge / longest
73	new_w, new_h = int(w * scale), int(h * scale)
74	img = img.resize((new_w, new_h), Image.LANCZOS)
75	out_name = os.path.splitext(fname)[0] + ".png"
76	img.save(os.path.join(output_dir, out_name), "PNG")
77	print(f"OK: {fname} → {out_name} ({img.size[0]}x{img.size[1]})")
78	except Exception as e:
79	print(f"FAIL: {fname} — {e}")
80	```
81
82	Rules:
83	- Max 2000px on the longest edge (configurable if user requests)
84	- Preserve aspect ratio
85	- Do NOT upscale — if already smaller than max, keep original dimensions
86	- Convert everything to PNG (RGBA mode for transparency support)
87
88	## Step 6: Remove Backgrounds
89
90	Check if `rembg` is installed. If not, install it:
91
92	```bash
93	pip3 install rembg onnxruntime
94	```
95
96	Then run background removal on all resized images → `nobg/`:
97
98	```python
99	from rembg import remove
100	from PIL import Image
101	import os, sys, io
102
103	input_dir = sys.argv[1] # resized/
104	output_dir = sys.argv[2] # nobg/
105
106	for fname in sorted(os.listdir(input_dir)):
107	if not fname.lower().endswith('.png'):
108	continue
109	try:
110	input_path = os.path.join(input_dir, fname)
111	with open(input_path, 'rb') as f:
112	input_data = f.read()
113	output_data = remove(input_data)
114	img = Image.open(io.BytesIO(output_data))
115	img.save(os.path.join(output_dir, fname), "PNG")
116	print(f"OK: {fname}")
117	except Exception as e:
118	print(f"FAIL: {fname} — {e}")
119	```
120
121	Note: The first run of rembg downloads the u2net model (~170MB). Warn the user

People who liked this also viewed

6 picks

Type