dot_claude/skills/x-post-craft/SKILL.md
Create viral X/Twitter posts based on the official X algorithm (Phoenix/Grok). Generates and improves posts optimized for maximum engagement. Use when (1) creating X posts for products, services, or information, (2) improving existing post drafts, (3) designing thread structures for long content.
npx skillsauth add paveg/dots x-post-craftInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Create high-engagement X/Twitter posts based on insights from X's public algorithm (Phoenix/Grok scoring system).
Reference: This skill is based on insights from x-algorithm - X's open-source recommendation algorithm.
X's algorithm predicts engagement probabilities and weights them into a final score:
| Positive Signal | Description | Weight |
|----------------|-------------|--------|
| dwell_time | Time spent reading | Very High |
| favorite | Like | High |
| retweet | Repost | High |
| reply | Reply | High |
| quote | Quote post | High |
| share / share_via_dm / copy_link | Share actions | Medium-High |
| follow_author | Follow | Medium |
| profile_click | Profile visit | Medium |
| vqv | Video quality view (30s+) | Video only |
| Negative Signal | Description | Impact |
|----------------|-------------|--------|
| not_interested | Not interested | Negative |
| block_author | Block | Strong Negative |
| mute_author | Mute | Strong Negative |
| report | Report | Strong Negative |
Same author's consecutive posts get score decay. → Quality over quantity - one great post beats multiple mediocre ones
Reaching non-followers requires higher scores. → Universal value spreads better than niche content
Collect from user:
Required:
- Topic (service/feature/news/insight)
- Core message (one sentence)
- Target audience
Optional:
- Tone (expert/casual/provocative/educational)
- Format (single/thread/quote RT)
- Visual assets (image/video)
- Target language for output
From collected info:
Minimum 3 patterns:
| Pattern | Characteristic | |---------|---------------| | A: Classic | Clear value, easy to read | | B: Hook-Heavy | Strong opening, curiosity gap | | C: Discussion-Starter | Strong opinion, invites replies |
For each variation:
Collect user preference, refine final version
[HOOK] 1-2 lines to stop the scroll
↓
[BODY] Deliver value
↓
[CTA] Call to action (optional)
| Pattern | Example | Goal | |---------|---------|------| | Number Impact | "5 things I learned after 10 years in tech" | Specificity builds trust | | Contrarian | "Everything you know about X is wrong" | Triggers discussion | | Empathy | "Struggling with X? Here's what worked" | Target identification | | Conclusion First | "Bottom line: You should do X" | Clarity | | Story | "Yesterday something happened that changed my view" | Curiosity |
| Format | Recommended Length | Reason | |--------|-------------------|--------| | Single (short) | 100-140 chars | Easy to RT | | Single (medium) | 200-280 chars | Value + completion | | Thread post 1 | ~140 chars | Hook focus | | Thread post 2+ | 200-280 chars | Expand content |
| Asset Type | Algorithm Effect | |------------|-----------------| | Infographic | dwell_time+, share+ | | Screenshot | Credibility, specificity | | Video (30s+) | vqv_score activation | | Image carousel | photo_expand+ |
1/N [HOOK] Why read this + "Thread below"
2/N [CONTEXT] Background or problem setup
3 to N-2/N [CORE] Main content (1 point per post)
N-1/N [SUMMARY] Key takeaways
N/N [CTA] Follow/RT/comment request
For improving existing drafts:
[ORIGINAL ANALYSIS]
- Strengths: ...
- Areas to improve: ...
[IMPROVED VERSION]
(improved post text)
[CHANGES EXPLAINED]
- Hook: [why changed]
- Body: [why changed]
- Algorithm view: [which signals improved]
## 📝 Post Variations
### A: Classic
(post text)
**Algorithm Analysis**: [why this structure scores well]
**Target Signals**: favorite, retweet, dwell_time
**Characters**: XXX
### B: Hook-Heavy
...
### C: Discussion-Starter
...
---
## 🎨 Visual Suggestions
### Image Option 1: [title]
- Content: [description]
- Expected effect: dwell_time+, share+
- Creation difficulty: Low/Medium/High
---
## 📊 Recommendation
**Most effective**: Variation [A/B/C]
**Reason**: [selection rationale]
## 🧵 Thread Structure
### 1/5 (Hook)
(hook post)
### 2/5 (Context)
(context post)
...
---
## 📊 Thread Design Breakdown
| Post | Role | Target Signals |
|------|------|---------------|
| 1/5 | Hook | click, dwell |
| 2/5 | Context | dwell_time |
| ... | ... | ... |
Use AskUserQuestion to collect:
Q1: What are you posting about?
- Product/service announcement
- News/trend commentary
- Knowledge/how-to sharing
- Personal experience/learning
- Other
Q2: Core message in one sentence
(free input)
Q3: Target audience?
- Industry peers/experts
- General users/consumers
- Specific community
- As broad as possible
Q4: What tone fits best?
- Expert/authoritative
- Friendly/approachable
- Educational/explanatory
- Provocative/debate-starting
- Humorous/light
Q5: Format?
- Single post (short)
- Single post (detailed)
- Thread
- Quote RT comment
Q6: Visual assets?
- Yes (will provide)
- No (suggest some)
- Undecided
Q7: Priority?
- Virality (RT focus)
- Credibility/expertise
- Discussion/comments
- Follower growth
Q8: Output language?
- English
- Japanese
- Other (specify)
Skill documentation: English Output language: User-specified (default: match user's input language)
When generating posts:
development
Iteratively improve agent-facing text instructions (skills, slash commands, task prompts, CLAUDE.md sections, code-generation prompts) by having a bias-free executor run them and evaluating from both sides (executor self-report + caller-side metrics). Repeat until improvement plateaus. Use immediately after creating or significantly revising a skill/prompt, or when unexpected agent behavior is suspected to stem from ambiguous instructions.
development
UI design quality standards and principles for frontend implementation and code review. Use when (1) implementing UI from design specs or mockups, (2) reviewing frontend/UI code, (3) creating new UI components, (4) building user interfaces for web or mobile apps. Complements frontend-design skills with quality enforcement.
development
Pre-submission review for iOS App Store. Scans the codebase for common rejection reasons and generates a pass/fail report with fixes. Use this skill when the user mentions App Store submission, review, release, 審査, 提出, or phrases like "ready to submit", "before submitting to Apple", "submission review", "rejection check". Also trigger when the user is preparing a TestFlight build for external review or discussing App Store rejection issues. Supports Swift/SwiftUI, UIKit, and React Native projects with Apple IAP and RevenueCat.
data-ai
Delegate an interrupt task to a background agent in an isolated worktree. The main session continues uninterrupted. The agent implements, commits, pushes, and creates a PR. Use when a quick fix or small task needs to happen without losing current context.