Name: doc-scan
Author: zengqing1218

Imported from zengqing1218/seryl (skills/seryl/doc-scan/SKILL.md). Install upstream with npx skills add zengqing1218/seryl --skill doc-scan. Copyright stays with the author (MIT).

Doc Scan（拍照转扫描件）

把手机拍的文档照片处理成扫描件效果（自动透视校正 + 自适应二值化/灰度增强）。

依赖（惰性安装，缺失时征得用户同意）

python -c "import cv2" 2>/dev/null || echo MISSING
uv pip install opencv-python-headless==5.0.0.93

用法（脚本: scripts/seryl_scan_enhance.py，仓库根目录下运行）

python scripts/seryl_scan_enhance.py photo.jpg scan.png              # 默认: 透视校正 + 黑白
python scripts/seryl_scan_enhance.py photo.jpg scan.png --mode gray  # 灰度增强（有照片/印章时）
python scripts/seryl_scan_enhance.py photo.jpg scan.png --no-warp    # 跳过透视校正

多页合成一个 PDF（Pillow）

python3 -c "
from PIL import Image
pages = [Image.open(p).convert('RGB') for p in ['scan1.png', 'scan2.png']]
pages[0].save('scanned.pdf', save_all=True, append_images=pages[1:], resolution=200)"

注意

透视校正靠边缘检测找文档四角：纸张与背景对比越明显越准；找不到四角会自动跳过校正（不报错）。
结果歪斜/裁错 → 加 --no-warp 重跑。
纯文字用默认黑白；含图片/图表/红章用 --mode gray。
扫描完要提取文字 → ocr-screenshot 或 ocr-and-documents 技能。

Install doc-scan and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

{ "ocm": "1", "id": "zengqing1218-seryl-doc-scan", "kind": "skill", "name": "doc-scan", "description": "拍照文档变扫描件：透视校正+黑白增强（opencv, 惰性安装）。", "publisher": "zengqing1218", "version": "1.0.0", "capabilities": { "domains": [ "general" ], "tags": [ "skill-md", "github" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "拍照文档变扫描件：透视校正+黑白增强（opencv, 惰性安装）。" ], "primary": false, "metadata": { "source": { "provider": "github", "repository": "https://github.com/zengqing1218/seryl", "path": "skills/seryl/doc-scan/SKILL.md", "ref": "90d53d4601b7ed84b867a71a22294e59e3e28756", "url": "https://github.com/zengqing1218/seryl/blob/90d53d4601b7ed84b867a71a22294e59e3e28756/skills/seryl/doc-scan/SKILL.md", "key": "zengqing1218/seryl/skills/seryl/doc-scan/SKILL.md" }, "license": "MIT" }, "instructions": "# Doc Scan（拍照转扫描件）\n\n把手机拍的文档照片处理成扫描件效果（自动透视校正 + 自适应二值化/灰度增强）。\n\n## 依赖（惰性安装，缺失时征得用户同意）\n\n```bash\npython -c \"import cv2\" 2>/dev/null || echo MISSING\nuv pip install opencv-python-headless==5.0.0.93\n```\n\n## 用法（脚本: scripts/seryl_scan_enhance.py，仓库根目录下运行）\n\n```bash\npython scripts/seryl_scan_enhance.py photo.jpg scan.png # 默认: 透视校正 + 黑白\npython scripts/seryl_scan_enhance.py photo.jpg scan.png --mode gray # 灰度增强（有照片/印章时）\npython scripts/seryl_scan_enhance.py photo.jpg scan.png --no-warp # 跳过透视校正\n```\n\n## 多页合成一个 PDF（Pillow）\n\n```bash\npython3 -c \"\nfrom PIL import Image\npages = [Image.open(p).co", "cost": { "context_tokens": 229 } }

doc-scan

About

Doc Scan（拍照转扫描件）

依赖（惰性安装，缺失时征得用户同意）

用法（脚本: scripts/seryl_scan_enhance.py，仓库根目录下运行）

多页合成一个 PDF（Pillow）

注意

Install doc-scan and it becomes one more candidate the router can pick - when it fits.

Use it

Manifest

Reviews