doc-review

Files (1)

• SKILL.md

SKILL.md (first lines)

Review a Word document and insert comments directly into the .docx file using python-docx and lxml XML manipulation.

Target file: $ARGUMENTS

## Modes

### 1. User-provided corrections
If the user provides a list of specific errors/corrections, insert a Word comment at each location in the document.

### 2. Full review (user says "full review" or similar)
Read the document and identify:
- Spelling errors and typos
- Grammar issues
- Inconsistent author name spellings
- Incomplete or malformed citations/references
- Template placeholder text that hasn't been filled in
- Tense inconsistencies
- Missing content (empty sections, placeholder tables)
- Formatting issues (inconsistent hyphenation, capitalization)

Then insert Word comments at each finding.

## How to insert Word comments

Use this Python approach with `python-docx` and `lxml`:

```python
from docx import Document
from docx.oxml.ns import qn
from docx.oxml import OxmlElement
from lxml import etree
from docx.opc.part import Part
from docx.opc.packuri import PackURI

# 1. Build the comments XML container
comments_xml = '''<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<w:comments xmlns:wpc="http://schemas.microsoft.com/office/word/2010/wordprocessingCanvas"
            xmlns:mc="http://schemas.openxmlformats.org/markup-compatibility/2006"
            xmlns:o="urn:schemas-microsoft-com:office:office"
            xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships"
            xmlns:m="http://schemas.openxmlformats.org/officeDocument/2006/math"
            xmlns:v="urn:schemas-microsoft-com:vml"
            xmlns:wp="http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing"
            xmlns:w10="urn:schemas-microsoft-com:office:word"
            xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main"
            xmlns:wne="http://schemas.microsoft.com/office/word/2006/wordml">
</w:comments>'''
comments_element = etree.fromstring(comments_xml.encode('utf-8'))

# 2. For each comment, add to the XML and mark the paragraph
def add_comment_to_xml(comments_element, comment_id, author, date_str, text):
    comment = etree.SubElement(comments_element, qn('w:comment'))
    comment.set(qn('w:id'), str(comment_id))
    comment.set(qn('w:author'), author)
    comment.set(qn('w:date'), date_str)
    p = etree.SubElement(comment, qn('w:p'))
    r = etree.SubElement(p, qn('w:r'))
    t = etree.SubElement(r, qn('w:t'))
    t.text = text
    t.set(qn('xml:space'), 'preserve')

def add_comment_markers(paragraph, search_text, comment_id):
    """Add commentRangeStart, commentRangeEnd, and commentReference to a paragraph."""
    # Try to find search_text in individual runs first
    found = False
    for run in paragraph.runs:
        if search_text.lower() in (run.text or "").lower():
            comment_start = OxmlElement('w:commentRangeStart')
            comment_start.set(qn('w:id'), str(comment_id))
            run.element.addprevious(comment_start)

            comment_end = OxmlElement('w:commentRangeEnd')
            comment_end.set(qn('w:id'), str(comment_id))
            run.element.addnext(comment_end)

            ref_run = OxmlElement('w:r')
            ref_rpr = OxmlElement('w:rPr')
            ref_style = OxmlElement('w:rStyle')
            ref_style.set(qn('w:val'), 'CommentReference')
            ref_rpr.append(ref_style)
            ref_run.append(ref_rpr)
            comment_ref = OxmlElement('w:commentReference')
            comment_ref.set(qn('w:id'), str(comment_id))
            ref_run.append(comment_ref)
            comment_end.addnext(ref_run)
            found = True
            break

    if not found:
        # Fallback: mark the whole paragraph
        p_element = paragraph._element
        comment_start = OxmlElement('w:commentRangeStart')
        comment_start.set(qn('w:id'), str(comment_id))
        p_element.insert(0, comment_start)

        comment_end = OxmlElement('w:commentRangeEnd')
        comment_end.set(qn('w:id'), str(comment_id))
        p_element.append(comment_end)

        ref_run = OxmlElement('w:r')
        ref_rpr = OxmlElement('w:rPr')
        ref_style = OxmlElement('w:rStyle')
        ref_style.set(qn('w:val'), 'CommentReference')
        ref_rpr.append(ref_style)
        ref_run.append(ref_rpr)
        comment_ref = OxmlElement('w:commentReference')
        comment_ref.set(qn('w:id'), str(comment_id))
        ref_run.append(comment_ref)
        p_element.append(ref_run)

# 3. After all comments are built, attach the comments part to the document
def attach_comments_part(doc, comments_element):
    comments_bytes = etree.tostring(comments_element, xml_declaration=True, encoding='UTF-8', standalone=True)
    doc_part = doc.part

    # Check for existing comments part
    existing = None
    for rel in doc_part.rels.values():
        if 'comments' in rel.reltype:

From jasonbhorne/claude-code-skills, MIT.

More from jasonbhorne/claude-code-skills