القائمة
Visually inspect a router bit image, determine the bounding box of the bit, crop to that region, and resize for the inventory page.
يُحدَّد Prompt الذي يراجع المصدر أولًا بشكل افتراضي. يمكنك التبديل إلى أمر مباشر أو تنزيل نسخة محلية.
اقرأ SKILL.md وأي ملفات مرافقة يعرضها SkillsMP قبل أن تقرر التثبيت.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
يتجاوز الأمر المباشر Prompt المخصّص للمراجعة. افحص المصدر قبل تشغيله.
npx skills add https://github.com/fjania/fjania.com --skill trim-bit-imageيبقى الأمر في سطر واحد. مرّر أفقيًا لمراجعته كاملًا قبل النسخ.
تفضّل نسخة محلية؟ نزّل الملفات المتاحة حاليًا لدى SkillsMP.
Add a router bit to the inventory by SKU — resolves vendor, fetches specs and image, creates a YAML data file.
Add a wood species sheet — researches the species online, creates a YAML data file, finds and downloads images, and builds the site.
Add a machine owner's manual — downloads the PDF, extracts text and images, translates to a readable HTML page with inline images, and builds the site.
استنادا إلى تصنيف SOC المهني
عرض SKILL.md
| name | trim-bit-image |
| description | Visually inspect a router bit image, determine the bounding box of the bit, crop to that region, and resize for the inventory page. |
| allowed-tools | Read, Bash |
| argument-hint | <image-path> |
Crop a router bit product image to just the bit itself, removing marketing text, captions, accessories, and excess whitespace.
Read the image at the path provided in $ARGUMENTS using the Read tool to visually inspect it.
Isolate objects using isolate-objects.py. This uses connected component analysis to find distinct objects on the white background and saves each as a separate image:
./workshop/router-bits/isolate-objects.py <image-path>
Output is one line per object (largest first):
<index> <outfile> <top>,<left>,<width>,<height> <pixel-area>
Use --max-y N to exclude caption text (default: 85% of image height). If the bit fills the full image height (the bearing or cutting head extends past 85%), pass --max-y 1000 or the bottom of the bit will be clipped.
Use --threshold N to adjust white detection (default: 252).
Fallback for lifestyle photos: If isolate-objects returns a single large object spanning most of the image, the photo has a scene (wood block, in-use shot, badge overlay) and the bit isn't separable on white. Recover it manually:
ImageDraw.rectangle(..., fill='white'), staying a few pixels outside the measured boundaries.isolate-objects.py on the cleaned image.Visually inspect each object using the Read tool. Identify which are router bits vs accessories (hex keys, collars, bearings, packaging). If there are multiple bits, each one will need to be processed.
For each router bit object, copy its isolated image file to the target path and resize with trim-image.sh:
cp <object-file> <image-path>
./workshop/router-bits/trim-image.sh <image-path>
Read the final image to verify — the bit should be centered, isolated on white, with no other objects or text visible. Check orientation against an existing bit of the same type (e.g. public/bit-images/US220618RO.jpg for roundovers): the convention is bearing-up / shank-down. If the source photo is shank-up, flip 180° with PIL (im.rotate(180)) before the final resize.
If isolation isn't clean (objects too close together, non-white background), adjust --threshold or --padding and re-run.