multi-platform-marketing - SKILL.md Agent Skill

name: multi-platform-marketing description: Automate multi-platform content distribution from a single source. Generate tailored marketing content for Twitter, LinkedIn, Instagram, Pinterest, and Email from one input (article, brief, or concept). Creates a "Content Fingerprint" (SSOT) with core message, keywords, and visual concepts, then adapts tone, format, and style for each platform. Use when distributing content across social platforms, launching campaigns, or repurposing content for multiple channels. Trigger phrases include "post to multiple platforms", "adapt for social media", "multi-platform campaign", "cross-platform content", or "distribute across platforms".

Multi-Platform Marketing

Eliminate the friction of multi-platform publishing. From a single input, automatically generate tailored assets for five platforms with consistent branding and precise tone-of-voice adaptation.

Core Concept: Content Fingerprint (SSOT)

Every multi-platform campaign starts with a Content Fingerprint - your Single Source of Truth containing:

Core Message: High-impact punchline (max 140 chars)
Key Keywords: 5 core hashtags for brand consistency
Visual Concepts: Descriptive keywords guiding image generation
Full Content: Original text for context
Content Type: Detected format (article, brief, announcement)

The fingerprint ensures all platform variations maintain brand consistency while respecting platform-specific norms.

Workflow

Step 1: Gather Input

Accept content in any of these formats:

Long-form article/blog post (500+ words)
Article (100-500 words)
Brief/concept (50-100 words)
Short announcement (<50 words)
Optional title for better core message extraction

Example prompts:

"Distribute this article across all platforms: [paste content]"
"Create multi-platform content for: We're launching a new AI collaboration tool"
"Adapt my blog post for social media: [paste content]"

Step 2: Generate Content Fingerprint

Run the fingerprint generation script:

python scripts/generate_fingerprint.py <content_file> [optional_title]
# OR
echo "Your content here" | python scripts/generate_fingerprint.py -

Output: JSON file containing the Content Fingerprint (SSOT)

What the script does:

Extracts the most impactful sentence as core message
Identifies 5 key topics/themes as hashtags
Detects sentiment and themes for visual concepts
Categorizes content type for platform adaptation

Step 3: Adapt for All Platforms

Run the platform adaptation script:

python scripts/adapt_content.py <fingerprint_json>
# OR
cat fingerprint.json | python scripts/adapt_content.py -

Output: JSON file containing adapted content for all 5 platforms:

Twitter: 1-3 tweet thread with hooks and hashtags
LinkedIn: Professional thought leadership post (authoritative tone)
Instagram: Emotional, emoji-rich caption with storytelling
Pinterest: Step-by-step, actionable pin description
Email: Compelling subject line + newsletter body

Each adaptation respects platform-specific:

Character limits
Tone and psychology
Hashtag strategy
Formatting conventions

See references/platform-specs.md for detailed platform requirements.

Step 4: Generate Visual Preview & Text Files

Create an HTML preview and individual text files:

# Generate HTML preview
python scripts/generate_html_preview.py <adapted_content_json> preview.html

# Extract individual platform files
# (You'll create these from the adapted JSON)

Outputs:

preview.html: Interactive preview showing all platforms
Individual .txt files for each platform (easy copy-paste)

Step 5: Generate Images (Optional)

Use the visual concepts from the fingerprint to generate platform-specific images:

# Use the generate-image skill with visual concepts
# Example: "professional, modern, technology, collaboration"

Integrate with the existing generate-image skill for actual image generation based on the visual concepts extracted in the fingerprint.

Step 6: Review & Publish

Open preview.html to see all platform variations side-by-side
Copy content from individual .txt files or use "Copy" buttons in HTML preview
Manually post to each platform (or use available automation)

Platform Adaptation Logic

Each platform receives content optimized for its unique psychology:

Twitter / X

Goal: Engagement and virality
Format: 1-3 tweet thread (280 chars each)
Tone: Conversational, hot takes, strong opinions
Hashtags: 1-2 per tweet, placed naturally
Strategy: Hook in tweet 1, insight in tweet 2, takeaway in tweet 3

Goal: Thought leadership
Format: Professional post (optimal 150-300 chars, max 3000)
Tone: Authoritative, data-driven, industry expertise
Hashtags: 3-5 industry-relevant hashtags at end
Strategy: Professional insight with personal touch, problem-solution-impact structure

Instagram

Goal: Emotional connection
Format: Caption (optimal 138-150 visible, max 2200)
Tone: Authentic, relatable, storytelling
Hashtags: 5-10 targeted hashtags (avoid generic spam)
Emojis: Liberal use (2-5+ per post)
Strategy: Personal story, emotional hook, community engagement CTA

Goal: Actionable education
Format: Pin description (optimal 200-300 chars, max 500)
Tone: Instructional, solution-oriented, DIY
Hashtags: 2-3 (less emphasis than other platforms)
Strategy: Numbered steps, keyword-rich, how-to focused

Email Newsletter

Goal: Click-through and engagement
Format: Subject (40-70 chars) + Body (50-125 words optimal)
Tone: Professional but approachable, conversational
Strategy: Curiosity-driven subject, scannable body with single CTA

Best Practices

Do

Start with the Content Fingerprint (SSOT) for brand consistency
Respect platform-specific character limits and tone
Adapt format and psychology for each platform
Use platform-specific hashtag strategies
Review the HTML preview before publishing
Generate images aligned with visual concepts

Don't

Copy-paste the same content across platforms
Use corporate jargon on Twitter or Instagram
Ignore character limits and formatting rules
Use 30 hashtags on Instagram (quality over quantity)
Write overly casual content for LinkedIn
Forget to hook readers in the first line

Visual Consistency

Use the visual concepts from fingerprint for image generation
Maintain brand colors and aesthetic across platforms
Adapt image dimensions per platform requirements
Ensure images support (not duplicate) text content

Reference Files

platform-specs.md: Detailed character limits, formatting rules, best practices, tone guidelines, and hashtag strategies for all platforms

Load these references when you need detailed platform specifications or edge case handling.

Scripts

All scripts accept input via file path or stdin (piped with -).

`generate_fingerprint.py`

Extracts Content Fingerprint (SSOT) from input content.

Usage:

python scripts/generate_fingerprint.py <content_file> [title]
echo "content" | python scripts/generate_fingerprint.py -

Output: JSON with core_message, key_keywords, visual_concepts, full_content, content_type

`adapt_content.py`

Transforms fingerprint into platform-specific content for all 5 platforms.

Usage:

python scripts/adapt_content.py <fingerprint_json>
cat fingerprint.json | python scripts/adapt_content.py -

Output: JSON with adapted content for twitter, linkedin, instagram, pinterest, email

`generate_html_preview.py`

Creates visual HTML preview showing all platforms side-by-side with copy buttons.

Usage:

python scripts/generate_html_preview.py <adapted_content_json> [output.html]

Output: Interactive HTML file with platform mockups

Example End-to-End

# 1. Create content fingerprint
echo "We're launching a revolutionary AI product that will change how teams collaborate. Our platform combines machine learning with intuitive design to make work more efficient." > content.txt
python scripts/generate_fingerprint.py content.txt "Launching Revolutionary AI Collaboration Tool" > fingerprint.json

# 2. Adapt for all platforms
python scripts/adapt_content.py fingerprint.json > adapted.json

# 3. Generate HTML preview
python scripts/generate_html_preview.py adapted.json preview.html

# 4. Open preview.html in browser to review
# 5. Copy content for each platform and publish

Integration with Other Skills

generate-image: Use visual concepts from fingerprint to generate images
bird: Auto-post Twitter threads (if configured)
send-email: Auto-send email newsletters (if configured)

Tips

Longer content (500+ words) produces richer adaptations
Include questions, stats, or numbers for better engagement hooks
Review and customize automated output for your brand voice
Test different content types to see what performs best per platform
Use the HTML preview to QA before publishing