Platform Voice Adapter

Best Practices:

• Lead with strongest point (hook in first 140 chars)
• Use line breaks for readability (max 2-3 lines per paragraph)
• Include 2-3 hashtags (more looks spammy)
• Add one emoji for visual break (don't overdo)
• Questions increase replies
• Numbered lists perform well in threads

Thread Specifications:

• Tweet 1: Hook that makes people want to click "Show More"
• Tweets 2-N: One clear point per tweet, numbered
• Final tweet: CTA (call to action) with link
• Optimal thread length: 5-10 tweets
• Each tweet should be standalone readable

Tone Spectrum:

• Technical topics: Approachable expert (not academic)
• Personal updates: Authentic, not overly casual
• Tips/tutorials: Direct instruction, use "you"

Formatting:

{Hook that creates curiosity}

{Main point in 1-2 lines}

{Supporting detail or benefit}

{2-3 hashtags}

LinkedIn

Character Limit: 3,000 characters (posts), 1,300 (optimal for full display) Optimal Length: 1,200-1,500 characters Voice: Professional, thoughtful, storytelling Format: Single post with visual hierarchy

Best Practices:

• Start with personal story or relatable scenario
• Use plenty of line breaks (LinkedIn rewards scannable content)
• Include data points and specific metrics
• Tag relevant people/companies (increases reach)
• End with engagement question
• 3-5 hashtags (LinkedIn surfaces hashtag content)
• Post during work hours (8am-6pm, Tue-Thu best)

Structure:

{Personal hook or story opening}

{Problem statement - relatable to audience}

{Your solution or insight}

Key takeaways:
• {Takeaway 1}
• {Takeaway 2}
• {Takeaway 3}

{Engagement question or CTA}

{Hashtags}

Tone Guidelines:

• More formal than Twitter, but not corporate stiff
• Show expertise through experience, not credentials
• Vulnerable moments increase engagement
• Avoid sales-y language (provide value first)
• Use "I" statements for personal posts, "you" for advice

Power Words for LinkedIn:

• Professional: strategy, framework, insight, approach, methodology
• Results: achieved, implemented, transformed, increased, optimized
• Learning: discovered, learned, realized, found, uncovered

Instagram

Caption Limit: 2,200 characters Optimal Length: 1,000-1,500 (full display without "more") Voice: Visual-first, aspirational, personal Format: Caption supports image/video

Best Practices:

• First line is critical (hook before "...more")
• Emojis are expected (use liberally but meaningfully)
• Stories for behind-the-scenes, feed for polished
• Hashtags: 5-10 relevant (can go in comments to keep caption clean)
• Tag location for local reach
• Ask questions to boost comments (algorithm reward)
• Carousel posts get highest engagement

Caption Structure:

{Hook in first line - visible before "more"}

{Story or context - 2-3 paragraphs}

{Value or lesson}

{CTA - tag friend, comment, save}

•••

{Hashtags - 5-10, can be in first comment instead}

Tone Spectrum:

• Personal brand: Authentic, conversational, emoji-friendly
• Educational: Simple language, visual metaphors
• Behind-the-scenes: Raw, unpolished is authentic

Emoji Usage:

• Use to create visual breaks (not every sentence)
• Replace common words (💰 for money, 🚀 for growth)
• Bullet points can be emojis (✨, 💡, 🔥)

Threads (Meta)

Character Limit: 500 characters per post Optimal Length: 300-400 characters Voice: Casual, authentic, conversational Format: Single post or thread

Best Practices:

• Most casual tone of all platforms
• Feels like texting with friends
• Share raw thoughts, not polished essays
• Lowercase acceptable in some contexts
• No hashtags culture (yet)
• Replies and quote posts highly visible
• "Hot takes" and opinions perform well

Structure:

{Casual conversational hook}

{Main thought - can be incomplete sentences}

{Maybe a follow-up observation}

Tone Guidelines:

• Write like you talk
• Shorter sentences
• More personal, less promotional
• Humor and personality shine
• Okay to be polarizing (not offensive)

Do's:

• ✅ Share opinions
• ✅ Ask open questions
• ✅ React to trends
• ✅ Thread longer thoughts

Don'ts:

• ❌ Over-formal language
• ❌ Corporate speak
• ❌ Excessive self-promotion
• ❌ Hashtag spam

TikTok (Text/Captions)

Caption Limit: 2,200 characters Optimal Length: 100-150 characters (people are watching video) Voice: Ultra-casual, trend-aware, punchy Format: Short caption supporting video

Best Practices:

• Caption is secondary to video
• Use trending sounds/formats
• Hashtags matter: 3-5 relevant + trending
• First 3 words hook (before "more")
• Direct, almost abrupt tone
• Lowercase trend is strong here
• Questions boost comments

Caption Formula:

{Hook in 3-5 words}

{Context if needed, ultra-brief}

{CTA or question}

{3-5 hashtags including trending}

Tone:

• Gen Z friendly (even if not your demo)
• Self-aware humor
• No corporate speak ever
• Embrace trends and memes
• Fast-paced, no fluff

Voice Adaptation Guidelines

Technical → Professional (Twitter → LinkedIn)

Twitter (Technical/Casual):

Just shipped a Claude Code agent that auto-generates content hooks using Kallaway frameworks 🔥

15 files, 3 hours, zero manual prompt engineering.

Thread on how I structured it 👇

LinkedIn (Professional/Storytelling):

I spent 3 hours yesterday building something that would've taken me 3 days last month.

The project: An autonomous content generation agent using Claude Code and proven viral hook frameworks.

The result: 15 Python files that transform raw project data into engagement-optimized social posts.

Here's what I learned about structuring AI agents for creative workflows:

• Separation of concerns beats monolithic prompts
• Framework-driven generation ensures consistency
• Progressive disclosure manages token budgets
• Event-driven architecture enables autonomy

The breakthrough wasn't the code—it was realizing that content frameworks are just structured prompts. Feed Claude the right pattern library, and it becomes a content strategist.

Building in public with #ClaudeCode. What content workflows are you automating?

#AI #Automation #ContentStrategy #BuildingInPublic

Professional → Casual (LinkedIn → Threads)

LinkedIn (Professional):

I discovered an interesting pattern while implementing context management for Claude Code agents.

By structuring documentation as hierarchical prompt libraries rather than static README files, we achieved 10x better context retention across sessions.

The key insight: Documentation isn't for humans anymore—it's for AI.

Threads (Casual):

wait

what if we stopped writing docs for humans and just wrote them as prompts for claude?

tested this in my personal-os project and it's actually wild how much better it works

like your readme.md becomes a literal instruction set that claude reads instead of you explaining the project every time

game changer honestly

Casual → Ultra-Casual (Instagram → TikTok)

Instagram:

POV: You just realized your documentation is actually prompt engineering 💡

I stopped writing README files for human readers and started writing them as instruction sets for Claude Code.

The result? Claude understands my projects instantly without me re-explaining the context every session.

Swipe for the 3-step framework I use →

#ClaudeCode #AI #Productivity #DeveloperTools #BuildingInPublic

TikTok:

docs are just prompts for claude

mind = blown 🤯

#claudecode #aitools #coding

Cross-Platform Adaptation Workflow

Step 1: Identify Source Content

Extract core elements:

• Main message: What's the one key point?
• Supporting points: What evidence/details exist?
• Call to action: What should audience do?
• Tone of source: Where is it on casual → professional spectrum?

Step 2: Select Target Platform(s)

Consider:

• Audience overlap: Is your LinkedIn audience the same as Twitter?
• Content type fit: Does this suit the platform? (Technical deep-dives great for LinkedIn, less so for TikTok)
• Format requirements: Can you support with visuals/video?

Step 3: Adapt Length

Expansion (Twitter → LinkedIn):

1. Add personal story context
2. Include data points and metrics
3. Expand single points into bullet lists
4. Add "key takeaways" section
5. Include engagement question

Compression (LinkedIn → Twitter):

1. Extract core insight
2. Remove story context (or save for thread)
3. Use power words to compact ideas
4. Focus on hook + one key point
5. Save supporting details for thread

Step 4: Adjust Tone

Professionalizing (casual → professional):

• Replace contractions (can't → cannot)
• Remove excessive emojis
• Add industry terminology
• Include credentials/data
• Structured formatting (bullets, numbers)

Casualizing (professional → casual):

• Use contractions
• Shorter sentences
• More conversational phrases ("honestly," "wait," "here's the thing")
• Add personality and humor
• Break grammar rules occasionally (lowercase, fragments)

Step 5: Format for Platform

Apply platform-specific:

• Line breaks: LinkedIn loves white space, Twitter needs compactness
• Emojis: Instagram/Threads expect them, LinkedIn uses sparingly
• Hashtags: LinkedIn 3-5, Instagram 5-10, Twitter 2-3, Threads minimal
• Structure: LinkedIn needs clear sections, Twitter needs flow

Step 6: Optimize for Engagement

Twitter: Ask questions, create threads, use polls LinkedIn: End with engagement question, tag relevant people Instagram: "Tag a friend who needs this," save/share prompts Threads: Controversial takes, open questions TikTok: Duet/stitch hooks, trending hashtags

Platform-Specific Templates

The Same Idea, 5 Platforms

Core Idea: "I built an agent that auto-generates viral hooks using Claude Code and the Kallaway framework. Saved 10 hours per week."

Twitter:

Just automated my content hook generation with Claude Code 🤖

Input: Project metadata
Output: Viral hooks using Kallaway frameworks

10 hours/week saved on ideation.

The secret? Treating frameworks as structured prompts.

#ClaudeCode #Automation

LinkedIn:

I automated the most time-consuming part of my content workflow this week.

The problem: I was spending 10+ hours weekly brainstorming social media hooks for my building-in-public updates.

The solution: An autonomous Claude Code agent that applies proven viral frameworks (Kallaway's contrarian, benefit-driven, transformation, and how-to patterns) to my daily project data.

How it works:
• Scans project activity (files changed, features shipped, problems solved)
• Matches patterns to appropriate hook frameworks
• Generates 3-5 hook variations per project update
• Outputs formatted for each platform (Twitter, LinkedIn, Threads)

The insight that made this work: Content frameworks are just structured prompts. Instead of manually applying "contrarian hook to project X," I feed Claude the framework pattern library and let it match context to template.

Week 1 results:
• 47 hooks generated from 12 project updates
• 10 hours saved
• Higher engagement (frameworks are proven patterns)
• Zero decision fatigue

Next step: Adding performance tracking so the agent learns which hook types perform best for which project categories.

Building in public with #ClaudeCode. What repetitive creative tasks are you automating?

#AI #ContentStrategy #Automation #BuildingInPublic #ProductivityHacks

Instagram:

POV: You just saved 10 hours per week on content ideation 💡

Swipe to see how I automated viral hook generation using Claude Code →

The framework:
1️⃣ Project metadata goes in
2️⃣ Claude applies Kallaway hook formulas
3️⃣ Viral hooks come out

No more staring at blank screens trying to make technical updates sound interesting.

The agent matches my project activity (new features, bugs fixed, lessons learned) to proven engagement patterns:
✨ Contrarian hooks for surprising insights
📈 Transformation hooks for before/after wins
🎯 How-to hooks for tutorials
💡 Benefit-driven hooks for value props

Week 1 = 47 hooks generated, zero creative block.

Tag someone who needs this ⬇️

#ClaudeCode #AI #ContentAutomation #BuildingInPublic #ProductivityHack #SoloFounder #CreatorEconomy

Threads:

i automated the "what should i post about" problem and it's honestly changed everything

built a claude code agent that reads my daily project activity and generates viral hooks using proven frameworks