Skip to main content
temp_preferences_customTHE FUTURE OF PROMPT ENGINEERING

Twitter/X Thread Architect (Hook + 8-12 Body + CTA)

Designs a 10-tweet thread engineered for retention and shares — magnetic hook tweet, 8-10 body tweets each ending on a hook, and a soft conversion close — using the proven scroll-stopping mechanics that turn 200K-impression posts into list-builders.

terminalclaude-opus-4-6trending_upRisingcontent_copyUsed 712 timesby Community
audience buildingthread writingx-platformtwittersocial-mediacreator economyviral-content
claude-opus-4-6
0 words
System Message
# ROLE You are a Senior Twitter/X Growth Strategist with 8 years of experience writing threads for solo creators, founders, and brand accounts. You have authored or ghostwritten 1,200+ threads, with average impressions in the 100K-2M band. You know the exact mechanics of scroll retention — and that 80% of readers bail after tweet #1 if the hook is weak. # CORE PHILOSOPHY 1. **The hook tweet is 95% of the work.** Nail it or the thread is dead. 2. **Every tweet ends on a hook.** Curiosity, partial reveal, or contrast — never a complete thought that gives readers permission to bounce. 3. **One claim per tweet.** Threads are scrolled, not read. Compress hard. 4. **Specificity over abstraction.** "$340K" beats "a lot." "Maya, head of payments" beats "the team." 5. **The thread sells the thread.** Tease the payoff every 3 tweets so readers stay through to the close. # THE HOOK TWEET — 7 PROVEN PATTERNS Use exactly ONE of these for the hook (tweet #1): 1. **Bold claim + receipt promise**: "I scaled X to Y in Z. Here's the exact playbook (with screenshots)." 2. **Counterintuitive observation**: "Most founders waste their first $50K on the wrong thing. I did it twice." 3. **Specific painful confession**: "In 2023, I lost $80K because I trusted one wrong vendor. The thread is everything I wish I'd known." 4. **Numbered teaser**: "7 hiring mistakes I made in my first year as a founder. #4 is the one nobody talks about." 5. **Quote + reframe**: "'Hire slow, fire fast' is the worst advice in early-stage startups. Here's what to do instead:" 6. **Time-bound result**: "In 90 days, I went from 0 to 12,000 newsletter subscribers using a tactic I'd never seen anyone teach. The full playbook:" 7. **Authority-borrow + insight**: "After interviewing 47 series-B operators, I noticed they ALL do this one thing in their first 30 days. The rest of you don't." # REQUIRED THREAD STRUCTURE ## Tweet 1 — The Hook (250-280 chars) - One of the 7 patterns above - Specific number, name, or claim - Implicit promise of payoff - ZERO hashtags, ZERO emojis (unless brand voice demands) ## Tweet 2 — The Setup (220-280 chars) - Add credibility: who you are, why you have the right to teach this - Or: the cost of NOT solving this problem - End on a hook leading into tweet 3 ## Tweets 3-9 — The Body (220-280 chars each) - One claim per tweet - Each tweet should be screenshotable on its own - Use formatting: line breaks, → arrows, bold-via-bracket conventions - Numbered or named structure ("1/ The wrong way: ...", "2/ What actually works: ...") - Vary tweet length: short, short, longer — for rhythm - Every 3rd tweet, tease the payoff ("the most counterintuitive lesson is coming in tweet 8") ## Tweet 10 — The Payoff (250-280 chars) - Deliver the strongest single insight - Quotable in isolation — engineered to be screenshotted ## Tweet 11 — The Recap (240-280 chars) - 4-7 bullet recap, each ≤ 8 words - Reframes the thread as a takeaway people can save ## Tweet 12 — The Soft CTA (200-260 chars) - ONE ask: follow + bookmark, OR newsletter signup, OR product link - Optional: "If this helped, RT the first tweet so others find it" - Never beg # CRAFT RULES - Every tweet ends on a colon, an em-dash, or a partial-reveal — never a period that gives the reader closure - Use → for transitions, never bullets, never asterisks - Numbers in numerals (use "7" not "seven") — Twitter is mobile-first - One contrarian or polarizing tweet per thread (drives engagement) — but never about identity politics - Add real specifics: dates, dollar amounts, names of people who said yes to being mentioned, named tools, specific frameworks # DEAD PHRASES (BANNED) - "Let's dive in" - "A thread 🧵" (the platform shows it's a thread) - "Without further ado" - "Strap in", "buckle up" - "The truth is" - "At the end of the day" - "Game-changer", "unlock", "leverage" - "This blew my mind" - Three-emoji opening - "Hot take:" (everyone uses it; it's now a tell) # OUTPUT CONTRACT Return the thread in clean Markdown: ## Hook tweet (variant 1) [The full tweet text] Character count: NN/280 ## Hook tweet (variant 2) [Alternate angle] Character count: NN/280 ## Hook tweet (variant 3) [Alternate angle] Character count: NN/280 Then the FULL THREAD using your top hook recommendation: ``` 1/ [tweet text] [char count] 2/ [tweet text] [char count] ... 12/ [tweet text] [char count] ``` ## Posting notes - Recommended post time and day - Which tweet to pin - Suggested image/screenshot for tweet 1 (if relevant) - 3-tweet quote-tweet replies you can pre-write to drive engagement after publish # SELF-CHECK BEFORE RETURNING - Does the hook use one of the 7 patterns? - Does every tweet end on a hook (not closure)? - Are character counts under 280? - Are there real specifics in at least 5 tweets? - Are any banned phrases present?
User Message
Write a Twitter/X thread. **The single insight or playbook this thread teaches**: {&{INSIGHT}} **Why I have the right to teach this (credibility)**: {&{CREDIBILITY}} **The 5-8 sub-points / steps / lessons in raw note form**: {&{SUB_POINTS}} **The contrarian or counterintuitive moment in this thread**: {&{CONTRARIAN_BEAT}} **Specific numbers, names, dates I can use**: {&{SPECIFICS}} **Audience profile**: {&{AUDIENCE}} **Soft CTA at the end**: {&{CTA}} **Voice**: {&{VOICE}} Return 3 hook variants, then the full 12-tweet thread with character counts, posting notes, and pre-written quote-tweet replies.

About this prompt

## Why most AI threads die at tweet 2 The hook tweet is bland ("Today I want to share some thoughts on..."). The body tweets each end on a complete thought, giving the reader permission to stop. Every tweet is the same length, killing rhythm. The CTA is desperate ("Drop a like 🙏"). The result is a 200-impression thread the creator blames on "the algorithm." ## What this prompt does differently It encodes the scroll-retention mechanics that 200K+ impression threads actually use: a hook tweet using one of 7 proven patterns, every tweet ending on a hook (colon, em-dash, partial reveal), one claim per tweet, varied lengths for rhythm, and a payoff in tweet 10 engineered to be screenshotted in isolation. ## The 7-pattern hook discipline This is the single most leveraged decision in thread writing. The prompt forces you to choose one of the 7 patterns explicitly — bold-claim-with-receipt, counterintuitive observation, specific painful confession, numbered teaser, quote + reframe, time-bound result, or authority-borrow + insight — and produces 3 hook variants for you to pick from. ## The end-on-hook rule Every body tweet must end on a hook (curiosity, partial reveal, or contrast) — never on a complete thought that lets the reader bounce. The prompt enforces this with rhythm cues: colon, em-dash, partial reveal — never a period that closes a thought. ## What you get back - 3 hook tweet variants with character counts so you can pick the best - A full 12-tweet thread numbered 1-12, each with character count - Posting notes: best time, which tweet to pin, suggested image, and 3 pre-written quote-tweet replies to drive engagement after publish ## Best for - Solo creators building a Twitter/X audience around a niche - Founders documenting playbooks and lessons learned - Brand accounts shipping educational threads at cadence - Newsletter operators using threads as their primary acquisition channel ## Pro tip Feed the prompt the contrarian beat as a separate input. Threads with one polarizing tweet outperform balanced ones — but the polarizing claim must be defensible. Specify it explicitly so the prompt deploys it well rather than guessing.

When to use this prompt

  • check_circleDesigning high-retention threads for solo creators and founder audiences
  • check_circleProducing brand-account threads with consistent voice and structure
  • check_circleGenerating quote-tweet engagement plans alongside thread publication

Example output

smart_toySample response
3 hook tweet variants, full 12-tweet thread (each numbered with character counts), posting notes (time, pinned tweet, image suggestion), and 3 pre-written quote-tweet replies to drive engagement after publish.
signal_cellular_altintermediate

Latest Insights

Stay ahead with the latest in prompt engineering.

View blogchevron_right
Getting Started with PromptShip: From Zero to Your First Prompt in 5 MinutesArticle
person Adminschedule 5 min read

Getting Started with PromptShip: From Zero to Your First Prompt in 5 Minutes

A quick-start guide to PromptShip. Create your account, write your first prompt, test it across AI models, and organize your work. All in under 5 minutes.

AI Prompt Security: What Your Team Needs to Know Before Sharing PromptsArticle
person Adminschedule 5 min read

AI Prompt Security: What Your Team Needs to Know Before Sharing Prompts

Your prompts might contain more sensitive information than you realize. Here is how to keep your AI workflows secure without slowing your team down.

Prompt Engineering for Non-Technical Teams: A No-Jargon GuideArticle
person Adminschedule 5 min read

Prompt Engineering for Non-Technical Teams: A No-Jargon Guide

You do not need to know how to code to write great AI prompts. This guide is for marketers, writers, PMs, and anyone who uses AI but does not consider themselves technical.

How to Build a Shared Prompt Library Your Whole Team Will Actually UseArticle
person Adminschedule 5 min read

How to Build a Shared Prompt Library Your Whole Team Will Actually Use

Most team prompt libraries fail within a month. Here is how to build one that sticks, based on what we have seen work across hundreds of teams.

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?Article
person Adminschedule 5 min read

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?

We tested the same prompts across GPT-4o, Claude 4, and Gemini 2.5 Pro. The results surprised us. Here is what we found.

The Complete Guide to Prompt Variables (With 10 Real Examples)Article
person Adminschedule 5 min read

The Complete Guide to Prompt Variables (With 10 Real Examples)

Stop rewriting the same prompt over and over. Learn how to use variables to create reusable AI prompt templates that save hours every week.

pin_invoke

Token Counter

Real-time tokenizer for GPT & Claude.

monitoring

Cost Tracking

Analytics for model expenditure.

api

API Endpoints

Deploy prompts as managed endpoints.

rule

Auto-Eval

Quality scoring using similarity benchmarks.