What problem does it solve? Merging chapters from multiple Word .docx files while preserving a designated template's formatting is error-prone: styleIds differ across same-source documents, VML images break, tracked changes leak into deliverables, section breaks get destroyed, and SEQ field numbering goes out of sequence. This Skill performs element-level surgery on word/document.xml using the format template as the ZIP base, so styles, numbering, headers, and section structure remain intact. ## Core Features & Use Cases - Chapter-level cross-document merge: Copy the template docx as base, slice body child elements by heading text and pStyle, deep-copy source blocks in, and remap pStyle/rStyle/tblStyle by style name rather than styleId. - Post-processing pipeline: Accept tracked changes (w:ins/w:del), remove dangling comment references, unify page size to A4 while preserving landscape sections, normalize captions, and renumber SEQ fields with a state machine that handles split instrText runs. - Specialized operations: Single-artifact extraction (large table plus caption into a template copy), supplementary image merging by media MD5 diff, in-place image replacement by overwriting media bytes only, cross-document table replacement, and SD design-number resequencing with traceability-matrix synchronization. - Use Case: A client template plus two content-source reports must be combined so chapters 1-4 come from document A, chapter 5.2 from document B, all formatted to the template, with tracked changes accepted and figure/table numbers resequenced before delivery. ## Quick Start Merge chapters 1 through 4 from the source report into the client template docx, keeping the template's formatting, accepting all tracked changes, and renumbering the figure and table captions.
