twitter-algorithm-optimizer

Analyze and optimize tweets for maximum reach using Twitter's open-source algorithm insights. Rewrite and edit user tweets to improve engagement and visibility based on how the recommendation system ranks content.

2,707 stars

Best use case

twitter-algorithm-optimizer is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

Analyze and optimize tweets for maximum reach using Twitter's open-source algorithm insights. Rewrite and edit user tweets to improve engagement and visibility based on how the recommendation system ranks content.

Teams using twitter-algorithm-optimizer should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

  • You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

  • You only need a quick one-off answer and do not need a reusable workflow.
  • You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/twitter-algorithm-optimizer/SKILL.md --create-dirs "https://raw.githubusercontent.com/davepoon/buildwithclaude/main/plugins/all-skills/skills/twitter-algorithm-optimizer/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/twitter-algorithm-optimizer/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How twitter-algorithm-optimizer Compares

Feature / Agenttwitter-algorithm-optimizerStandard Approach
Platform SupportNot specifiedLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

Analyze and optimize tweets for maximum reach using Twitter's open-source algorithm insights. Rewrite and edit user tweets to improve engagement and visibility based on how the recommendation system ranks content.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

Related Guides

SKILL.md Source

# Twitter Algorithm Optimizer

## When to Use This Skill

Use this skill when you need to:
- **Optimize tweet drafts** for maximum reach and engagement
- **Understand why** a tweet might not perform well algorithmically
- **Rewrite tweets** to align with Twitter's ranking mechanisms
- **Improve content strategy** based on the actual ranking algorithms
- **Debug underperforming content** and increase visibility
- **Maximize engagement signals** that Twitter's algorithms track

## What This Skill Does

1. **Analyzes tweets** against Twitter's core recommendation algorithms
2. **Identifies optimization opportunities** based on engagement signals
3. **Rewrites and edits tweets** to improve algorithmic ranking
4. **Explains the "why"** behind recommendations using algorithm insights
5. **Applies Real-graph, SimClusters, and TwHIN principles** to content strategy
6. **Provides engagement-boosting tactics** grounded in Twitter's actual systems

## How It Works: Twitter's Algorithm Architecture

Twitter's recommendation system uses multiple interconnected models:

### Core Ranking Models

**Real-graph**: Predicts interaction likelihood between users
- Determines if your followers will engage with your content
- Affects how widely Twitter shows your tweet to others
- Key signal: Will followers like, reply, or retweet this?

**SimClusters**: Community detection with sparse embeddings
- Identifies communities of users with similar interests
- Determines if your tweet resonates within specific communities
- Key strategy: Make content that appeals to tight communities who will engage

**TwHIN**: Knowledge graph embeddings for users and posts
- Maps relationships between users and content topics
- Helps Twitter understand if your tweet fits your follower interests
- Key strategy: Stay in your niche or clearly signal topic shifts

**Tweepcred**: User reputation/authority scoring
- Higher-credibility users get more distribution
- Your past engagement history affects current tweet reach
- Key strategy: Build reputation through consistent engagement

### Engagement Signals Tracked

Twitter's **Unified User Actions** service tracks both explicit and implicit signals:

**Explicit Signals** (high weight):
- Likes (direct positive signal)
- Replies (indicates valuable content worth discussing)
- Retweets (strongest signal - users want to share it)
- Quote tweets (engaged discussion)

**Implicit Signals** (also weighted):
- Profile visits (curiosity about the author)
- Clicks/link clicks (content deemed useful enough to explore)
- Time spent (users reading/considering your tweet)
- Saves/bookmarks (plan to return later)

**Negative Signals**:
- Block/report (Twitter penalizes this heavily)
- Mute/unfollow (person doesn't want your content)
- Skip/scroll past quickly (low engagement)

### The Feed Generation Process

Your tweet reaches users through this pipeline:

1. **Candidate Retrieval** - Multiple sources find candidate tweets:
   - Search Index (relevant keyword matches)
   - UTEG (timeline engagement graph - following relationships)
   - Tweet-mixer (trending/viral content)

2. **Ranking** - ML models rank candidates by predicted engagement:
   - Will THIS user engage with THIS tweet?
   - How quickly will engagement happen?
   - Will it spread to non-followers?

3. **Filtering** - Remove blocked content, apply preferences

4. **Delivery** - Show ranked feed to user

## Optimization Strategies Based on Algorithm Insights

### 1. Maximize Real-graph (Follower Engagement)

**Strategy**: Make content your followers WILL engage with

- **Know your audience**: Reference topics they care about
- **Ask questions**: Direct questions get more replies than statements
- **Create controversy (safely)**: Debate attracts engagement (but avoid blocks/reports)
- **Tag related creators**: Increases visibility through networks
- **Post when followers are active**: Better early engagement means better ranking

**Example Optimization**:
- ❌ "I think climate policy is important"
- ✅ "Hot take: Current climate policy ignores nuclear energy. Thoughts?" (triggers replies)

### 2. Leverage SimClusters (Community Resonance)

**Strategy**: Find and serve tight communities deeply interested in your topic

- **Pick ONE clear topic**: Don't confuse the algorithm with mixed messages
- **Use community language**: Reference shared memes, inside jokes, terminology
- **Provide value to the niche**: Be genuinely useful to that specific community
- **Encourage community-to-community sharing**: Quotes that spark discussion
- **Build in your lane**: Consistency helps algorithm understand your topic

**Example Optimization**:
- ❌ "I use many programming languages"
- ✅ "Rust's ownership system is the most underrated feature. Here's why..." (targets specific dev community)

### 3. Improve TwHIN Mapping (Content-User Fit)

**Strategy**: Make your content clearly relevant to your established identity

- **Signal your expertise**: Lead with domain knowledge
- **Consistency matters**: Stay in your lanes (or clearly announce a new direction)
- **Use specific terminology**: Helps algorithm categorize you correctly
- **Reference your past wins**: "Following up on my tweet about X..."
- **Build topical authority**: Multiple tweets on same topic strengthen the connection

**Example Optimization**:
- ❌ "I like lots of things" (vague, confuses algorithm)
- ✅ "My 3rd consecutive framework review as a full-stack engineer" (establishes authority)

### 4. Boost Tweepcred (Authority/Credibility)

**Strategy**: Build reputation through engagement consistency

- **Reply to top creators**: Interaction with high-credibility accounts boosts visibility
- **Quote interesting tweets**: Adds value and signals engagement
- **Avoid engagement bait**: Doesn't build real credibility
- **Be consistent**: Regular quality posting beats sporadic viral attempts
- **Engage deeply**: Quality replies and discussions matter more than volume

**Example Optimization**:
- ❌ "RETWEET IF..." (engagement bait, damages credibility over time)
- ✅ "Thoughtful critique of the approach in [linked tweet]" (builds authority)

### 5. Maximize Engagement Signals

**Explicit Signal Triggers**:

**For Likes**:
- Novel insights or memorable phrasing
- Validation of audience beliefs
- Useful/actionable information
- Strong opinions with supporting evidence

**For Replies**:
- Ask a direct question
- Create a debate
- Request opinions
- Share incomplete thoughts (invites completion)

**For Retweets**:
- Useful information people want to share
- Representational value (tweet speaks for them)
- Entertainment that entertains their followers
- Information advantage (breaking news first)

**For Bookmarks/Saves**:
- Tutorials or how-tos
- Data/statistics they'll reference later
- Inspiration or motivation
- Jokes/entertainment they'll want to see again

**Example Optimization**:
- ❌ "Check out this tool" (passive)
- ✅ "This tool saved me 5 hours this week. Here's how to set it up..." (actionable, retweet-worthy)

### 6. Prevent Negative Signals

**Avoid**:
- Inflammatory content likely to be reported
- Targeted harassment (gets algorithmic penalty)
- Misleading/false claims (damages credibility)
- Off-brand pivots (confuses the algorithm)
- Reply-guy syndrome (too many low-value replies)

## How to Optimize Your Tweets

### Step 1: Identify the Core Message
- What's the single most important thing this tweet communicates?
- Who should care about this?
- What action/engagement do you want?

### Step 2: Map to Algorithm Strategy
- Which Real-graph follower segment will engage? (Followers who care about X)
- Which SimCluster community? (Niche interested in Y)
- How does this fit your TwHIN identity? (Your established expertise)
- Does this boost or hurt Tweepcred?

### Step 3: Optimize for Signals
- Does it trigger replies? (Ask a question, create debate)
- Is it retweet-worthy? (Usefulness, entertainment, representational value)
- Will followers like it? (Novel, validating, actionable)
- Could it go viral? (Community resonance + network effects)

### Step 4: Check Against Negatives
- Any blocks/reports risk?
- Any confusion about your identity?
- Any engagement bait that damages credibility?
- Any inflammatory language that hurts Tweepcred?

## Example Optimizations

### Example 1: Developer Tweet

**Original**:
> "I fixed a bug today"

**Algorithm Analysis**:
- No clear audience - too generic
- No engagement signals - statements don't trigger replies
- No Real-graph trigger - followers won't engage strongly
- No SimCluster resonance - could apply to any developer

**Optimized**:
> "Spent 2 hours debugging, turned out I was missing one semicolon. The best part? The linter didn't catch it.
>
> What's your most embarrassing bug? Drop it in replies 👇"

**Why It Works**:
- SimCluster trigger: Specific developer community
- Real-graph trigger: Direct question invites replies
- Tweepcred: Relatable vulnerability builds connection
- Engagement: Likely replies (others share embarrassing bugs)

### Example 2: Product Launch Tweet

**Original**:
> "We launched a new feature today. Check it out."

**Algorithm Analysis**:
- Passive voice - doesn't indicate impact
- No specific benefit - followers don't know why to care
- No community resonance - generic
- Engagement bait risk if it feels like self-promotion

**Optimized**:
> "Spent 6 months on the one feature our users asked for most: export to PDF.
>
> 10x improvement in report generation time. Already live.
>
> What export format do you want next?"

**Why It Works**:
- Real-graph: Followers in your product space will engage
- Specificity: "PDF export" + "10x improvement" triggers bookmarks (useful info)
- Question: Ends with engagement trigger
- Authority: You spent 6 months (shows credibility)
- SimCluster: Product management/SaaS community resonates

### Example 3: Opinion Tweet

**Original**:
> "I think remote work is better than office work"

**Algorithm Analysis**:
- Vague opinion - doesn't invite engagement
- Could be debated either way - no clear position
- No Real-graph hooks - followers unclear if they should care
- Generic topic - dilutes your personal brand

**Optimized**:
> "Hot take: remote work works great for async tasks but kills creative collaboration.
>
> We're now hybrid: deep focus days remote, collab days in office.
>
> What's your team's balance? Genuinely curious what works."

**Why It Works**:
- Clear position: Not absolutes, nuanced stance
- Debate trigger: "Hot take" signals discussion opportunity
- Question: Direct engagement request
- Real-graph: Followers in your industry will have opinions
- SimCluster: CTOs, team leads, engineering managers will relate
- Tweepcred: Nuanced thinking builds authority

## Best Practices for Algorithm Optimization

1. **Quality Over Virality**: Consistent engagement from your community beats occasional viral moments
2. **Community First**: Deep resonance with 100 engaged followers beats shallow reach to 10,000
3. **Authenticity Matters**: The algorithm rewards genuine engagement, not manipulation
4. **Timing Helps**: Engage early when tweet is fresh (first hour critical)
5. **Build Threads**: Threaded tweets often get more engagement than single tweets
6. **Follow Up**: Reply to replies quickly - Twitter's algorithm favors active conversation
7. **Avoid Spam**: Engagement pods and bots hurt long-term credibility
8. **Track Your Performance**: Notice what YOUR audience engages with and iterate

## Common Pitfalls to Avoid

- **Generic statements**: Doesn't trigger algorithm (too vague)
- **Pure engagement bait**: "Like if you agree" - hurts credibility long-term
- **Unclear audience**: Who should care? If unclear, algorithm won't push it far
- **Off-brand pivots**: Confuses algorithm about your identity
- **Over-frequency**: Spamming hurts engagement rate metrics
- **Toxicity**: Blocks/reports heavily penalize future reach
- **No calls to action**: Passive tweets underperform

## When to Ask for Algorithm Optimization

Use this skill when:
- You've drafted a tweet and want to maximize reach
- A tweet underperformed and you want to understand why
- You're launching important content and want algorithm advantage
- You're building audience in a specific niche
- You want to become known for something specific
- You're debugging inconsistent engagement rates

Use Claude without this skill for:
- General writing and grammar fixes
- Tone adjustments not related to algorithm
- Off-Twitter content (LinkedIn, Medium, blogs, etc.)
- Personal conversations and casual tweets

Related Skills

x-twitter-scraper

2707
from davepoon/buildwithclaude

X (Twitter) data extraction and monitoring via Xquik: tweet search, user lookup, follower extraction, giveaway draws, trending topics, account monitoring with webhooks, reply/retweet/quote extraction, community and Space data, follow checks. 22 MCP tools + REST API.

twitter-automation

2707
from davepoon/buildwithclaude

Automate Twitter/X tasks via Rube MCP (Composio): posts, search, users, bookmarks, lists, media. Always search tools first for current schemas.

soft-screening-startup

2707
from davepoon/buildwithclaude

Activate for ANY startup evaluation, investment screening, or company assessment. Triggers include: "evaluate this startup", "screen this company", "should I invest in X", "is this a good investment", "what do you think about this company", "review this startup", "score this company", "rate this pitch", "assess this founder", "quick take on X", "is X worth investing in", "pass or decline on X", "what's your verdict on X", "first look at this company", "quick screen on X", "what's your take on this founder", "is this fundable", "would a VC invest in this". Also triggers when a user pastes a company description, funding ask, or founder background and asks for an opinion. Works on claude.ai and Claude Code. For hard-mode deterministic scoring with Python audit trail, use /venture-capital-intelligence:hard-screening-startup.

market-size

2707
from davepoon/buildwithclaude

Run TAM/SAM/SOM market sizing with top-down and bottom-up methods, competitive landscape, and tech stack analysis. Triggered by: "/venture-capital-intelligence:market-size", "size this market", "what is the TAM for X", "market sizing analysis", "competitive landscape for X", "who are the competitors", "TAM SAM SOM for X", "market opportunity analysis", "how big is this market", "is this market big enough", "what's the addressable market", "total addressable market for X", "how large is the opportunity", "market research for X", "how saturated is this market", "market size estimate", "go-to-market sizing", "what is the serviceable market". Claude Code only. Requires Python 3.x. Uses web search for market data.

hard-screening-startup

2707
from davepoon/buildwithclaude

Deterministic Python-scored startup screening with full audit trail. Use when you need a reproducible, weighted-score verdict on a startup — not just a qualitative opinion. Triggered by: "/venture-capital-intelligence:hard-screening-startup", "hard screen this startup", "run a hard screen on X", "score this startup with Python", "give me an auditable screen", "run a scored evaluation on X", "give me a weighted score for this startup", "screen with numbers", "objective startup score", "reproducible screen", "investment scorecard for X", "score this company out of 100", "run the full screen on X". Claude Code only. Requires Python 3.x. For conversational soft-mode screening, use /venture-capital-intelligence:soft-screening-startup.

fund-operations

2707
from davepoon/buildwithclaude

Compute fund KPIs (TVPI, DPI, IRR, MOIC), model carried interest and management fees, and generate LP quarterly update narratives. Triggered by: "/venture-capital-intelligence:fund-operations", "calculate fund KPIs", "what is my fund TVPI", "IRR calculation", "compute MOIC", "LP report", "quarterly update draft", "carried interest calculation", "management fee calculation", "fund performance report", "write my LP update", "how is my fund performing", "what is my DPI", "fund returns analysis", "model my carry", "how much carry do I earn", "portfolio performance summary", "generate investor update". Claude Code only. Requires Python 3.x.

financial-model

2707
from davepoon/buildwithclaude

Run deterministic financial models for startup valuation and SaaS health analysis. Triggered by: "/venture-capital-intelligence:financial-model", "run a financial model on X", "DCF this company", "model the financials", "calculate runway", "what is the valuation", "SaaS metrics model", "LTV CAC analysis", "unit economics", "burn rate analysis", "comparable valuation", "how long is my runway", "what's my burn multiple", "revenue projection for X", "model the ARR growth", "what is the pre-money valuation", "comps analysis", "NRR and churn model", "how healthy are these SaaS metrics". Claude Code only. Requires Python 3.x. Accepts user-supplied numbers or searches for publicly available data.

explain-equity-terms

2707
from davepoon/buildwithclaude

Activate for ANY equity, legal, or term sheet question related to startup investing or fundraising. Triggers include: "what is a SAFE", "explain this term sheet", "what does pro-rata mean", "what is liquidation preference", "explain anti-dilution", "ISO vs NSO", "what is a 83(b) election", "what is carried interest", "explain drag-along", "what is a valuation cap", "what does MFN mean", "explain convertible note vs SAFE", "what is a down round", "explain vesting cliff", "what does fully diluted mean", "term sheet question", "equity question", "what does this clause mean". Also triggers when a user pastes legal text from a term sheet, SAFE, or subscription agreement and asks what it means. Works on claude.ai and Claude Code.

deal-sourcing-signals

2707
from davepoon/buildwithclaude

Scan a company or sector for deal-sourcing signals across 6 dimensions. Triggered by: "/venture-capital-intelligence:deal-sourcing-signals", "scan signals for X", "what signals is X showing", "deal sourcing scan", "hiring signals for X", "is X raising soon", "monitor this company", "company signal scan", "sourcing brief for X", "what is X up to", "is X growing", "track this company", "deal signal report for X", "is this company fundraising", "what are the momentum signals for X", "find signals on X", "is X worth tracking". Claude Code only. Requires Python 3.x. Uses web search for live signal data.

cap-table-waterfall

2707
from davepoon/buildwithclaude

Model cap table dilution, SAFE conversion, and exit waterfall across scenarios. Triggered by: "/venture-capital-intelligence:cap-table-waterfall", "model my cap table", "simulate dilution", "SAFE conversion math", "exit waterfall", "how much do I own after Series A", "liquidation waterfall", "cap table scenario", "what happens to equity at exit", "model the waterfall", "how much equity do I have left", "what is my ownership after funding", "run dilution scenarios", "model a new round", "what happens at acquisition", "cap table after SAFE conversion", "pari passu waterfall", "preference stack analysis". Claude Code only. Requires Python 3.x.

analyze-pitch-deck

2707
from davepoon/buildwithclaude

Activate for ANY pitch deck analysis, feedback, or review request. Triggers include: "analyze this deck", "review my pitch deck", "critique my pitch", "feedback on my slides", "is my deck investor ready", "what's wrong with my pitch", "how would a VC react to this deck", "score my pitch deck", "rate my slides", "improve my deck", "what slides am I missing", "is this pitch compelling". Also triggers when a user pastes slide content, describes their deck structure, or shares a company narrative and asks for investor feedback. Works on claude.ai and Claude Code.

public-plugin-builder

2707
from davepoon/buildwithclaude

Activate when the user wants to build a Claude plugin, create a Claude skill, make a Claude agent, structure a Claude Code plugin, says "build a plugin", "create a skill", "new claude skill", "new agent", "help me make a plugin", "plugin builder", "claude plugin helper", "how do I build a Claude skill", "I want to create a Claude plugin", "plugin building", or asks how to structure a Claude Code plugin or publish to the Claude marketplace. Works on both claude.ai (generates files as code blocks) and Claude Code (writes and pushes files).