菜单
Analyze Elon Musk's repost patterns on X and generate tweets optimized for getting reposted by @elonmusk. Use when users want to craft viral AI/tech product launches targeting Musk's attention, go viral on X/Twitter, or optimize tweets for maximum reach.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
CMO brain + execution for AI agents. Brand analysis, competitive intelligence, GEO/AEO visibility audits, AI search optimization, on-page SEO, keyword research, Reddit sentiment monitoring, Reddit content strategy, X/Twitter viral optimization, web search, webpage analysis, content marketing, and growth strategy. Requires API key from app.karis.im/agent.
Audit your brand's AI search visibility and website GEO readiness. Supports two modes — visibility audit (are you being mentioned by AI?) and site audit (is your site optimized for AI?). Use when users ask about AI search performance, GEO/AEO scores, brand mentions in ChatGPT/Perplexity/Claude, AI crawler accessibility, llms.txt, or content optimization for generative engines.
Analyze a brand and recommend high-leverage growth actions. Use when users provide a brand or URL for analysis, ask what marketing actions to take, want to understand their competitive position, or need help increasing visibility and traffic. Also use before running other Karis skills to establish brand context.
Optimize a webpage for traditional search engines (Google, Bing). Covers keyword research, competitor page analysis, on-page SEO audit with scoring, content optimization, and schema markup recommendations. Use when users ask about SEO scores, title tags, meta descriptions, keyword optimization, heading structure, internal linking, or structured data for search engines.
Generate Reddit growth content including viral posts, comment strategies, and engagement plans. Use when founders or marketers want to create Reddit content for brand awareness, product launches, or community building. Works by calling Karis CLI which provides content strategy, post drafts, and engagement recommendations.
Analyze Reddit discussions for brand intelligence, competitor insights, and community sentiment. Use when users want to understand Reddit conversations about their brand, track competitor mentions, discover community pain points, analyze sentiment trends, or monitor brand health on Reddit. Works by calling Karis CLI which connects to Reddit data via Karis API.
| name | elonmusk-repost |
| description | Analyze Elon Musk's repost patterns on X and generate tweets optimized for getting reposted by @elonmusk. Use when users want to craft viral AI/tech product launches targeting Musk's attention, go viral on X/Twitter, or optimize tweets for maximum reach. |
| metadata | {"version":"1.2.0","author":"karis-ai"} |
Analyze what Elon Musk reposts on X and generate tweets optimized for his engagement.
| User Says | Use This Skill |
|---|---|
| "Help me go viral on X" | elonmusk-repost |
| "Optimize my tweet for Musk" | elonmusk-repost |
| "Write a product launch tweet" | elonmusk-repost |
| "Score my tweet for repost potential" | elonmusk-repost |
| "Get Elon to retweet this" | elonmusk-repost |
npx @karis-ai/cli) for deeper brand contextkaris setup)When invoked, go straight to scoring and tweet generation in a single response. Do NOT ask clarifying questions unless critical info is truly missing (no product name AND no description at all). Extract everything you can from the user's message — product name, what it does, media, metrics, draft text — and immediately produce the full output (score + tweets + strategy).
Speed is critical. The user expects one message in, one message out. Never split the analysis across multiple turns.
If the user provides a draft tweet, score THAT draft and generate improved variants.
If the user has Karis CLI configured, pull brand context first:
npx @karis-ai/cli brand show
Use brand profile data (competitors, audience, differentiators) to sharpen the tweet angle. If no brand profile exists, proceed without it.
Task Progress:
- [ ] Step 1: Extract product info (from user message or Karis brand profile)
- [ ] Step 2: Score against Musk repost patterns
- [ ] Step 3: Generate optimized tweets
- [ ] Step 4: Provide posting strategy
Evaluate the user's content against these proven patterns:
| Pattern | Weight | Description |
|---|---|---|
| AI Visual Demo | 10 | Visually stunning AI-generated output (images, video, 3D, games) |
| AI Agent | 9 | Autonomous AI that completes complex tasks end-to-end |
| "Wow Factor" | 9 | Instantly impressive, no explanation needed |
| Robotics / Physical AI | 9 | Humanoid robots, embodied AI, real-world autonomous systems |
| Video / Multimodal Generation | 8 | AI-generated video, voice, music, or cross-modal output |
| Playable/Interactive | 8 | Users can try it immediately (game, tool, demo) |
| Grok/xAI Adjacent | 8 | Content that relates to or complements Grok capabilities |
| Open-Source AI | 7 | Open-weight models, open-source tools (Musk pushed Grok open-source) |
| Underdog / Unexpected Origin | 7 | Surprising team origin — small team, non-obvious country, solo dev |
| Creator Tool | 7 | Empowers individuals to create professional-grade output |
| Speed/Scale | 6 | "X million users in Y hours" type virality metrics |
| Government Efficiency / DOGE | 6 | Tools that cut bureaucratic waste or align with efficiency themes |
| Meme Potential | 6 | Content that can be remixed into memes |
| Anti-Establishment | 5 | Challenges incumbents (Adobe, Google, OpenAI, etc.) |
| Space / Mars | 5 | Space exploration, Mars colonization, rocket tech |
For each pattern the content matches:
match_score(i) = relevance(0-10) × weight(i)
Repost Score = Σ match_score(i) / Σ weight(i) for ALL patterns × 100
relevance = how strongly the content matches that pattern (0 = no match, 10 = perfect match)weight = the number in the table above| Product | Score | Key Patterns Hit |
|---|---|---|
| Lovart (AI Design Agent) | 81 | AI Visual Demo, AI Agent, Underdog, Creator Tool, Speed/Scale, Wow, Anti-Establishment |
| Loopit (AI Game Gen) | 79 | Playable, AI Visual Demo, Underdog, Creator Tool, Wow, Grok Adjacent |
| Midjourney | 78 | AI Visual Demo, Wow, Creator Tool, Anti-Establishment, Speed/Scale, Meme |
| Suno (AI Music) | 76 | AI Visual Demo, Wow, Multimodal, Creator Tool, Playable, Meme |
| Manus (AI Agent) | 74 | AI Agent, Underdog, Wow, Speed/Scale |
| DeepSeek | 73 | Open-Source, Underdog, Anti-Establishment, Wow, Speed/Scale |
| Cursor (AI Code Editor) | 72 | AI Agent, Wow, Creator Tool, Anti-Establishment, Speed/Scale, Grok Adjacent |
Products scoring 70+ all share: strong visual demo OR autonomous agent + wow factor. Products below 50 typically lack visual impact.
Avoid these — they virtually guarantee being ignored:
Based on analysis of content Musk engages with, use this structure:
[Hook - 1 line, max impact] + [Visual media] + [Optional: metric/claim]
| Window | Time (PT) | Notes |
|---|---|---|
| Prime | 10 PM - 2 AM | Musk's late-night scroll; highest engagement probability |
| Secondary | 6 PM - 10 PM | Evening wind-down; decent reach |
| Weekend | Sunday 8 PM - 1 AM | Variable but Sunday evening is consistently strong |
| Avoid | 6 AM - 12 PM | Low engagement; posts get buried before Musk is active |
Template A: Product Launch
[Product Name] — [bold claim in <=8 words]
[1-2 sentence description of what it does]
[Attach: 30-60s demo video or GIF]
Example:
Lovart — the world's first AI design agent.
Give it a brief. It delivers agency-grade creative in minutes.
20,000 people queued on launch day.
[45s demo video attached]
Template B: Milestone/Traction
[Metric] in [timeframe].
[Product Name] lets you [core value prop].
[Attach: demo video showing the "wow" moment]
Example:
1M songs created in 72 hours.
Suno turns a one-line prompt into a full, radio-quality track.
[30s audio/video demo attached]
Template C: Challenge the Status Quo
[Incumbent] charges $X/mo for [thing].
[Product Name] does it with AI for [fraction of price / free].
[Attach: side-by-side comparison video]
Example:
Adobe charges $55/mo for design tools most people never learn.
Lovart does it with one sentence for $9/mo.
[side-by-side comparison: 2 hours in Photoshop vs 30 seconds in Lovart]
Template D: Pure Demo (Highest Conversion)
[No text or minimal text — just the demo video/GIF]
[Reply to own tweet with product name + link]
Example:
[60s screen recording: type a prompt → watch a playable 3D game appear in real-time]
Reply: "This is Loopit. Try it -> loopit.ai"
Template E: Reply to Musk's Tweet
[Monitor Musk's tweets on a relevant topic]
[Reply with your demo + one punchy line — no @tag needed, you're already in his thread]
Example:
Musk tweets: "AI is getting insane"
You reply: "Made a game in 10 seconds with pure AI. No code." + [15s GIF]
This is often the highest-conversion path — you're already in his notification stream.
For deeper content optimization, use Karis CLI:
npx @karis-ai/cli chat --skill elonmusk-repost "Write a viral X/Twitter launch tweet for [product]. Context:
- Product: [description]
- Target: Elon Musk repost
- Include: hook, wow factor, demo description
- Keep under 50 words
- Use product launch template"
Produce ALL of this in a single response — no follow-up turns:
## Repost Score: X/100
| Pattern | Relevance | Score |
|---------|-----------|-------|
| [only patterns scoring >0] | X/10 | XX |
## Tweet Variants
### Variant 1 (Template X)
[tweet text under 50 words]
Media: [what to attach]
### Variant 2 (Template X)
[tweet text under 50 words]
Media: [what to attach]
### Variant 3 (Template X)
[tweet text under 50 words]
Media: [what to attach]
## Posting Strategy
- Best time: [specific window]
- Media: [what to prepare]
- Thread strategy: [if applicable]
## How to Increase Score (+X points)
1. [actionable change] → +X points
2. [actionable change] → +X points
3. [actionable change] → +X points
When Musk reposts or engages, his style is:
Content that elicits these minimal-effort responses has the highest repost probability.
Don't: beg-tag Musk, delete/repost repeatedly, or spam across accounts.