ima-claude 2.9.0 → 2.10.0

package/dist/cli.js

@@ -11,7 +11,7 @@ var HOOKS_DIR = join(CLAUDE_DIR, "hooks");
 var COMMANDS_DIR = join(CLAUDE_DIR, "commands");
 var RULES_DIR = join(CLAUDE_DIR, "rules");
 var SETTINGS_FILE = join(CLAUDE_DIR, "settings.json");
-var VERSION = "2.9.0";
+var VERSION = "2.10.0";
 var colors = {
   reset: "\x1B[0m",
   bright: "\x1B[1m",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "ima-claude",
-  "version": "2.9.0",
+  "version": "2.10.0",
   "description": "IMA's AI coding agent skills - FP patterns, architecture guidance, and team standards. Supports Claude Code, Junie CLI, and more.",
   "type": "module",
   "scripts": {

package/plugins/ima-claude/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "ima-claude",
-  "version": "2.9.0",
+  "version": "2.10.0",
   "description": "IMA's Claude Code skills for functional programming, architecture, and team standards. 47 skills, 24 hooks, default persona, 3-tier memory system.",
   "author": {
     "name": "IMA",

package/plugins/ima-claude/skills/ima-doc2pdf/SKILL.md

@@ -0,0 +1,242 @@
+---
+name: ima-doc2pdf
+description: >-
+  Convert DOCX content into branded IMA PDF documents using ReportLab with Lato
+  typography, navy headings, justified body text, running footers, and embedded
+  images. Produces branded PDF documents with content pages. Generates a placeholder
+  cover that can be replaced by ima-cover-creator for production output. Use when: converting a Word doc to branded PDF,
+  creating PDF content pages for Canva import, generating an IMA branded document PDF,
+  or when the user says "convert this docx to PDF," "branded PDF," "content pages,"
+  "PDF for Canva," or "doc to PDF." Also triggers on: "make a PDF from this Word
+  file," "export to PDF," "generate branded PDF," "IMA branded document PDF."
+  Always load ima-brand alongside for color/typography authority.
+---
+
+# IMA DOCX → Branded PDF
+
+Extracts content from any Word document and generates a branded IMA PDF with
+Lato typography, navy/gold colors, and IMA layout standards. Works with guides,
+reports, white papers, and other IMA documents. Outputs content pages only
+(no cover page) — the cover is handled by `ima-cover-creator` and merged via pypdf.
+
+---
+
+## Why ReportLab for Content
+
+Content pages are pure document flow: headings, body paragraphs, bullet lists,
+inline bold/italic, embedded images, and running footers. ReportLab handles all of
+this with proper Lato font registration, automatic page breaks, and precise
+typographic control. No coordinate math needed — the flow engine does the work.
+
+---
+
+## Quick Start
+
+### 1. Install dependencies (once)
+
+```bash
+pip install python-docx reportlab Pillow pypdf --break-system-packages
+```
+
+### 2. Generate content PDF
+
+```bash
+python3 scripts/generate_pdf.py path/to/document.docx --out content.pdf
+```
+
+### 3. Merge with cover (from ima-cover-creator)
+
+```python
+from pypdf import PdfReader, PdfWriter
+
+cover = PdfReader("cover.pdf")
+content = PdfReader("content.pdf")
+
+writer = PdfWriter()
+writer.add_page(cover.pages[0])
+
+# Skip the first 2 pages (ReportLab's placeholder cover + overflow)
+for page in content.pages[2:]:
+    writer.add_page(page)
+
+with open("final.pdf", "wb") as f:
+    writer.write(f)
+```
+
+---
+
+## Pipeline
+
+```
+DOCX
+  ↓  extract_docx.py (text, structure, metadata)
+  ↓  generate_pdf.py (ReportLab → branded PDF)
+  ↓
+content.pdf (N pages, no cover)
+  +
+cover.pdf (from ima-cover-creator)
+  ↓  pypdf merge
+  ↓
+final.pdf → Canva import
+```
+
+---
+
+## Typography Spec (Canva-confirmed)
+
+All values confirmed from Canva design data. Font: Lato (Google Fonts).
+
+### Headings
+
+| Element | Size | Weight | Color | Align |
+|---------|------|--------|-------|-------|
+| Section heading (h2) | 15pt | Bold | #00066F | Center |
+| Sub-heading (h3) | 13pt | Bold | #00066F | Left |
+| Intro heading | 15pt | Bold | #00066F | Center |
+
+### Body Text
+
+| Element | Size | Weight | Color | Align |
+|---------|------|--------|-------|-------|
+| Body paragraph | 12pt | Regular | #000000 | Justify |
+| Body bold inline | 12pt | Bold | #000000 | Justify |
+| Body bold navy | 12pt | Bold | #00066F | Justify |
+| Bullet item | 12pt | Regular | #000000 | Left |
+| Bullet marker | — | — | #00066F | — |
+
+### Other Elements
+
+| Element | Size | Weight | Color |
+|---------|------|--------|-------|
+| Footer | 10pt | Regular | #666666 |
+| Reference entry | 8pt | Regular | #333333 |
+| Reference heading | 13pt | Bold | #00066F |
+| Q&A question | 12pt | Bold | #00066F |
+| Q&A answer | 12pt | Regular | #000000 |
+| Warning box | 12pt | Bold | #FFFFFF on #00066F bg |
+
+### Page Setup
+
+| Property | Value |
+|----------|-------|
+| Page size | US Letter (8.5 × 11 in) |
+| Margins | 0.5 in all sides |
+| Body width | 7.5 in |
+| Footer height | 0.4 in from bottom |
+
+---
+
+## Content Extraction
+
+The `extract_docx.py` script classifies each Word paragraph into typed blocks:
+
+| Type | Description |
+|------|-------------|
+| `h1` | Top-level heading (title) |
+| `h2` | Section heading |
+| `h3` | Sub-heading |
+| `heading_bold` | All-bold paragraph (inline heading) |
+| `body` | Regular paragraph |
+| `bullet` | List item |
+| `author` | Author name |
+| `date` | Date string |
+| `disclaimer` | Disclaimer text |
+| `warning` | Warning box content |
+| `question` | Q&A question |
+| `answer_start` | Q&A answer (YES/NO prefix) |
+| `reference` | Numbered citation |
+| `ref_heading` | "References" heading |
+| `figure_caption` | Figure/table caption |
+| `page_break` | Hard page break |
+
+Each block includes `runs` with per-run bold/italic flags for inline formatting.
+
+---
+
+## Image Handling
+
+The script extracts embedded DOCX images via `python-docx`:
+
+1. Reads all image relationships from the DOCX package
+2. Maps paragraph indices to embedded image positions
+3. Writes images to temp files
+4. Inserts ReportLab `Image` flowables at the correct positions
+5. Scales to fit within `page_width - 2 × margin`
+
+Images that appear between text paragraphs (image-only paragraphs) are also caught
+and appended after all text content.
+
+---
+
+## Cover Page Behavior
+
+The ReportLab script generates a **placeholder cover** (navy background with title
+text) as pages 1-2 of its output. This exists so the script works standalone, but
+when pairing with `ima-cover-creator`, **skip the first 2 pages** during merge.
+
+To check which pages to skip:
+```python
+from pypdf import PdfReader
+r = PdfReader("content.pdf")
+for i in range(min(3, len(r.pages))):
+    text = r.pages[i].extract_text()[:100]
+    print(f"Page {i}: {text}")
+```
+
+The first content page typically starts with "Introduction" or a section heading.
+
+---
+
+## Scripts
+
+| Script | Purpose |
+|--------|---------|
+| `generate_pdf.py` | Main: DOCX → branded PDF via ReportLab |
+| `extract_docx.py` | Extracts structured content from Word documents |
+| `docx_utils.py` | Shared utilities for DOCX parsing |
+
+---
+
+## Fonts
+
+Lato TTF files are auto-downloaded from Google Fonts on first run into the `fonts/`
+directory (which is git-ignored). The font family (Regular, Bold, Italic, BoldItalic)
+is registered with ReportLab so that `<b>` and `<i>` markup works in Paragraph objects.
+
+If the fonts are already present, the download is skipped.
+
+---
+
+## Customization
+
+### Adjusting the footer
+
+The footer shows the document title and date. To customize:
+```python
+# In generate_pdf.py, the footer text is built from:
+title_short = (title[:60] + "...") if len(title) > 60 else title
+footer_text = f"{title_short} ({clean_date})"
+```
+
+### Adding new paragraph types
+
+1. Add a classifier rule in `extract_docx.py`
+2. Add a handler in `block_to_flowables()` in `generate_pdf.py`
+3. Create a ReportLab `ParagraphStyle` in `build_styles()`
+
+---
+
+## Relationship to Other Skills
+
+| Skill | Role |
+|-------|------|
+| **ima-cover-creator** | Generates branded cover page (PPTX → PDF) |
+| **ima-cancer-care-guides** | Full pipeline including Markdown source and Canva API mapping |
+| **ima-brand** | Source of truth for colors, typography, voice |
+
+**Typical workflow:**
+```
+ima-cover-creator  →  cover.pdf (1 page)
+ima-doc2pdf        →  content.pdf (skip first 2 pages)
+pypdf merge        →  final.pdf → Canva import
+```

package/plugins/ima-claude/skills/ima-doc2pdf/references/formatting-spec.md

@@ -0,0 +1,88 @@
+# Content Page Formatting Spec
+
+Values confirmed from Canva design data (Cancer Drug Resistance Guide, March 2026).
+Canva internal units × 0.75 = points.
+
+## Page Setup
+
+| Property | Value |
+|----------|-------|
+| Page size | US Letter (8.5 × 11 in / 612 × 792 pt) |
+| Margins | 0.5 in (36 pt) all sides |
+| Body text width | 7.5 in (540 pt) |
+| Footer position | 0.4 in from bottom |
+
+## Color Palette
+
+| Name | Hex | Usage |
+|------|-----|-------|
+| Trustworthy Indigo | #00066F | Headings, bullet markers, warning bg |
+| Body Black | #000000 | Body text (pure black, not #1A1A1A) |
+| Dark Gray | #333333 | Reference entries |
+| Gray Text | #666666 | Footer |
+| Light Gray | #CCCCCC | Footer rule |
+| White | #FFFFFF | Warning text |
+| Vital Gold | #FFCC00 | Warning emphasis |
+
+## Section Heading (h2)
+
+- Font: Lato 15pt Bold
+- Color: #00066F (navy)
+- Alignment: Center
+- Spacing: 16pt above, 5pt below
+
+## Sub-heading (h3)
+
+- Font: Lato 13pt Bold
+- Color: #00066F (navy)
+- Alignment: Left
+- Spacing: 10pt above, 3pt below
+
+## Body Paragraph
+
+- Font: Lato 12pt Regular
+- Color: #000000
+- Alignment: Justify
+- Line height: 1.4 (14.5pt leading)
+- Spacing: 6pt below
+
+## Bullet Lists
+
+- Font: Lato 12pt Regular
+- Color: #000000
+- Bullet marker color: #00066F (navy)
+- Indent: 18pt
+- Spacing: 1pt above, 1pt below each item
+
+## Bold Inline Variants
+
+- **Body bold** (#000000): Drug names, emphasis
+- **Body bold navy** (#00066F): Inline sub-headings
+
+## Warning Box
+
+- Background: #00066F (navy) — note: ReportLab uses solid, Canva uses gradient
+- Text: Lato 12pt Bold, #FFFFFF
+- Emphasis: Vital Gold #FFCC00
+- Alignment: Center
+- Padding: 10pt all sides
+
+## Footer
+
+- Font: Lato 10pt
+- Color: #666666
+- Alignment: Center
+- Content: "{Document Title} ({Date})"
+- Rule above: 0.5pt, #CCCCCC
+
+## References
+
+- Heading: Lato 13pt Bold, #00066F
+- Entry: Lato 8pt Regular, #333333
+- Hanging indent: 14pt
+
+## Q&A
+
+- Question: Lato 12pt Bold, #00066F
+- Answer: Lato 12pt Regular, #000000
+- YES/NO prefix: Bold

package/plugins/ima-claude/skills/ima-doc2pdf/scripts/docx_utils.py

@@ -0,0 +1,21 @@
+"""
+Shared utilities for Word document processing across ima-cancer-care-guides scripts.
+"""
+
+from docx.oxml.ns import qn
+
+
+def has_page_break(para):
+    """Detect a hard page break in a Word paragraph.
+
+    Checks both explicit w:br type=page runs and section-level page breaks
+    (w:pPr/w:sectPr), which appear on the last paragraph of a section.
+    """
+    for run in para.runs:
+        for br in run._element.findall(qn('w:br')):
+            if br.get(qn('w:type')) == 'page':
+                return True
+    pPr = para._element.find(qn('w:pPr'))
+    if pPr is not None and pPr.find(qn('w:sectPr')) is not None:
+        return True
+    return False