slide-stylist

name	slide-stylist
description	Derive a complete StyleGuide through collaborative design exploration with the Speaker, working within brand constraints when a BrandProfile exists.
argument-hint	["--deck-dir PATH"]
allowed-tools	Bash(python *), Read, Glob, Write

/slide-stylist

Derive a complete StyleGuide (palette, typography, layout templates, image style tokens) through collaborative exploration with the Speaker.

Prerequisites

• ./tmp/deck/talk-brief.json must exist
• ./tmp/deck/brand-profile.json may exist (produced by brand-manager)

Usage

Invoked by the Deck Conductor after brand-manager. Can also be invoked directly:

/slide-stylist
/slide-stylist --deck-dir ./tmp/deck

What It Does

Step 1: Read Inputs

Read ./tmp/deck/talk-brief.json for topic, audience, tone, and preferences. Read ./tmp/deck/brand-profile.json if it exists for brand constraints.

Determine the design mode:

• BrandProfile with strict compliance: Palette and fonts locked. Only propose accent, image mood, layout preferences.
• BrandProfile with guided compliance: Brand provides foundation. Propose variations.
• No BrandProfile: Full creative latitude.

Step 2: Palette Direction

With strict brand: Skip -- use BrandProfile palette directly. Tell the Speaker:

"Using your brand palette (strict compliance). Primary: #{primary}, Secondary: #{secondary}."

With guided brand: Propose 2-3 variations:

"Your brand palette starts with Primary: #{primary}. Here are three accent directions:

1. Split-complementary warm: Accent #{accent1} -- energetic, draws attention to CTAs
2. Analogous cool: Accent #{accent2} -- cohesive, professional feel
3. Triadic balanced: Accent #{accent3} -- vibrant, good for data-heavy decks

Which direction? (Or suggest your own)"

No brand: Propose 2-3 complete palettes based on topic + tone:

"For a {tone} talk about {topic}, here are three palette directions:

1. {name1}: Primary #{p1}, Secondary #{s1}, Accent #{a1} -- {description}
2. {name2}: Primary #{p2}, Secondary #{s2}, Accent #{a2} -- {description}
3. {name3}: Primary #{p3}, Secondary #{s3}, Accent #{a3} -- {description}

Which direction? (Or describe what you're looking for)"

After Speaker selects, derive the full palette (background, text colours, chart series) ensuring all contrast ratios meet thresholds.

Step 3: Font Pairing

With brand font preference: Use brand font as heading, select best body complement:

"Using {brand_font} for headings (per brand). For body text, I recommend {body_font} -- {rationale}. Does that work?"

Without brand font: Select from the 10 scored pairings based on tone and topic:

Tone	Recommended Pairing
technical	Inter (single family) -- #4
professional	Montserrat + Open Sans -- #1
conversational	Montserrat + Open Sans -- #1
inspirational	Playfair Display + Source Sans Pro -- #2
provocative	Oswald + Merriweather -- #8
storytelling	Raleway + Lato -- #5

Topic keyword overrides:

• AI/ML/data/engineering/developer -> Inter or Work Sans + IBM Plex Sans
• startup/pitch/product -> Poppins + Roboto
• design/creative/art -> Raleway + Lato or Space Grotesk + DM Sans
• research/academic/science -> Fira Sans + Source Serif Pro

Present the recommendation:

"For a {tone} talk about {topic}, I recommend:

• Heading: {heading_font}
• Body: {body_font}
• Code: {mono_font}

{brief rationale}. Does this feel right?"

Step 4: Image Style Tokens

Derive from tone + preferences.style + palette:

"For the visual style of generated images, I suggest:

• Mood: {mood}
• Colour direction: {color_direction}
• Style: {style_modifiers}

This means images will have a {description}. Does that match your vision?"

Mood mapping:

• professional -> "clean and authoritative"
• conversational -> "friendly and approachable"
• technical -> "precise and structured"
• inspirational -> "bold and aspirational"
• provocative -> "edgy and thought-provoking"
• storytelling -> "warm and narrative"

Style modifier mapping from preferences.style:

• minimalist -> ["clean lines", "ample white space", "simple geometry", "flat colour"]
• image-rich -> ["high detail", "photographic quality", "rich textures", "dramatic lighting"]
• data-heavy -> ["clean infographic style", "structured layouts", "subtle gradients"]
• diagram-heavy -> ["precise lines", "clear labels", "structured flow", "technical illustration"]
• corporate -> ["professional photography", "polished", "brand-consistent", "executive"]
• creative -> ["artistic", "bold composition", "experimental", "vibrant textures"]

Step 5: Produce StyleGuide

Assemble the final StyleGuide with:

• Agreed palette (all 10 colour values)
• Agreed typography (heading_font, body_font, mono_font, heading_sizes, body_size, caption_size, line_spacing)
• All 12 layout templates (these are fixed engineering constants, not design choices):

Slide Type	text_zone (x,y,w,h)	image_zone (x,y,w,h)	background_treatment
title	(1.0, 1.5, 8.0, 2.5)	(0, 0, 10.0, 5.625)	image_bleed
section_divider	(1.0, 1.5, 8.0, 2.5)	null	solid_dark
content	(0.5, 1.0, 5.5, 4.0)	(6.5, 0.5, 3.0, 4.625)	solid_light
two_column	(0.5, 1.0, 4.0, 4.0)	(5.5, 1.0, 4.0, 4.0)	solid_light
image_feature	(0.5, 3.5, 4.0, 1.5)	(0, 0, 10.0, 3.2)	image_bleed
data_chart	(0.5, 0.5, 9.0, 1.0)	(0.5, 1.6, 9.0, 3.5)	solid_light
stat_callout	(1.5, 1.0, 7.0, 3.5)	null	solid_dark
quote	(1.5, 1.0, 7.0, 3.0)	null	solid_dark
icon_grid	(0.5, 0.5, 9.0, 1.0)	(0.5, 1.6, 9.0, 3.5)	solid_light
diagram	(0.5, 0.5, 9.0, 1.0)	(0.5, 1.8, 9.0, 3.3)	solid_light
closing	(1.0, 1.0, 8.0, 3.5)	null	gradient
blank_visual	null	(0, 0, 10.0, 5.625)	image_bleed

• Agreed image style tokens (mood, color_direction, style_modifiers[])

Font sizes (fixed defaults from Research #10):

• title_slide: 44, section_divider: 36, slide_heading: 28, subheading: 20
• body_size: 16, caption_size: 12, line_spacing: 1.4

Layout dimensions (fixed):

• slide_width_inches: 10, slide_height_inches: 5.625, margin_inches: 0.5

Plugin Setup

PLUGIN_ROOT=$(python3 -c "
from pathlib import Path
import sys, os
if os.environ.get('JACK_TAR_DECKHAND_ROOT'):
    print(os.environ['JACK_TAR_DECKHAND_ROOT']); sys.exit()
home = Path.home()
for base in [home / '.claude' / 'plugins' / 'cache']:
    for p in base.rglob('jack-tar-deckhand/.claude-plugin/plugin.json'):
        print(str(p.parent.parent)); sys.exit()
dev = Path.cwd() / 'plugins' / 'jack-tar-deckhand'
if dev.exists():
    print(str(dev)); sys.exit()
print('NOT_FOUND')
" 2>/dev/null)
if [ -z "$PLUGIN_ROOT" ] || [ "$PLUGIN_ROOT" = "NOT_FOUND" ]; then echo "ERROR: jack-tar-deckhand not found" && exit 1; fi

Step 6: Validate and Write

Validate the StyleGuide before writing:

PYTHONPATH="$PLUGIN_ROOT" python3 -c "
import json
from src.style_validation import validate_style_guide, check_palette_contrast, check_completeness
sg = json.load(open('./tmp/deck/style-guide.json'))
schema_errors = validate_style_guide(sg)
contrast_issues = check_palette_contrast(sg.get('palette', {}))
completeness_issues = check_completeness(sg)
for e in schema_errors: print(f'SCHEMA: {e}')
for e in contrast_issues: print(f'CONTRAST: {e}')
for e in completeness_issues: print(f'COMPLETENESS: {e}')
if not schema_errors and not contrast_issues and not completeness_issues:
    print('StyleGuide validates successfully')
"

If validation fails, fix the issues and re-validate. Write the validated StyleGuide to ./tmp/deck/style-guide.json.

Output

./tmp/deck/style-guide.json

Design Rules

• All hex values are 6 characters WITHOUT the # prefix (PptxGenJS convention)
• Left-align body text, centre only titles
• Target 7:1+ contrast ratio for all text/background pairs (projection readability)
• Never skip the Speaker collaboration steps -- always propose and get approval
• Layout templates are fixed constants -- do not ask the Speaker about coordinates
• Font sizes are fixed defaults -- do not ask the Speaker about point sizes