Agent Skills
› codewhale-hq/Codewhale
› pdf
处理PDF文件的技能，支持读取、提取、分割、合并、旋转、水印、表单填写、OCR及创建等操作。提供工具选择建议，强调保留原文档、验证输出内容及OCR准确性。
Trigger Scenarios
需要读取或解析PDF内容
需要对PDF进行编辑如合并、拆分或加水印
需要从扫描版PDF中提取文本
Install
npx skills add codewhale-hq/Codewhale --skill pdf -g -y
SKILL.md
Frontmatter
{
"name": "pdf",
"description": "Read, extract, split, merge, rotate, watermark, fill, OCR, or create PDF files with verification of page counts and text extraction."
}
Use this skill for any task where a PDF is the primary input or output.
Workflow
- Identify the PDF operation: read, extract, OCR, split, merge, rotate, watermark, redact, fill forms, encrypt/decrypt, or create.
- Preserve originals. Write outputs with explicit names.
- Use the most reliable available tool:
pdftotext,pdfinfo,qpdf, ormutoolthroughbashwhen installed (the built-inreadtool does not extract text from a PDF)- Python libraries such as
pypdf,pdfplumber,PyMuPDF, orreportlabwhen available - OCR tools only for scanned pages
- For extraction, report page coverage and note when layout, tables, or OCR quality may affect accuracy.
- For generated or modified PDFs, verify page count, text extraction where possible, and file size. For redaction, confirm removed text is not extractable from the output.
Ask before installing dependencies or running OCR over large documents. Do not represent a visually scanned PDF as fully accurate text unless OCR quality has been checked.
Version History
-
94130d9
Current 2026-09-27 21:49
修正工具引用，将模型指向可见的read/bash命令，移除对内置File工具提取PDF文本的错误描述。
-
0fe366b
2026-08-16 09:03
将过时的内部工具名（如read_file）替换为正确的模型可见名称（File action: read），修复提示词中的错误引导。
- b0e4926 2026-07-24 17:42
