Agent skill (LLM usage)

docx_plus ships an Agent Skill — an LLM-facing guide that teaches a coding agent how to drive the library correctly: the units, the "mark_fields_dirty before save" rule for fields, style IDs vs. names, toggle semantics, and the per-module APIs. If you use Claude Code (or any agent that reads skill files) to generate .docx automation, point it at this skill instead of hand-feeding API snippets.

What's in it

The skill lives at skills/docx-plus/ in the repository, using the standard umbrella + progressive-disclosure layout: one SKILL.md entry point that an agent loads first, plus topic reference files it pulls in on demand.

File	Covers
SKILL.md	Entry point — mental model, cross-cutting conventions, capability map
reference/forms.md	FormBuilder; read / set / clear control values; document protection
reference/styles.md	Cascade inspection + provenance; create / modify / apply / ensure / remap; theme
reference/publishing.md	TOC, captions, table of figures, footnotes, endnotes, bookmarks, cross-references, fields
reference/layout.md	Columns, section breaks, even/odd headers, line numbering, page borders
reference/comments.md	Anchored comments (add / edit / delete / read)

Using it

• Claude Code — copy (or symlink) the skills/docx-plus directory into your project's .claude/skills/ folder, or into your personal ~/.claude/skills/. The agent discovers it through the name: / description: frontmatter and loads it when a task matches.
• Any other LLM / RAG pipeline — feed SKILL.md as context; it points to the reference files, which the agent reads on demand. Everything is plain Markdown with runnable Python.

Accuracy

Every code snippet in the skill imports only public symbols and is verified to run end to end against the current release (v0.2.0). The skill mirrors the same public surface documented in the API Index and the per-module Reference pages — it's the agent-facing complement to those human-facing docs, not a separate source of truth.