A data-driven skill for scoring and optimizing tweets based on X’s actual open-source ranking algorithm.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install tweet-composer
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install tweet-composer using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Tweet Composer?
Tweet Composer is a specialized utility designed for creators and marketers who want to move beyond guesswork and align their social media strategy with technical reality. Unlike generic writing tools, this skill leverages insights derived from X's Phoenix ranking pipeline to score drafts based on 19 predicted engagement actions. By integrating Tweet Composer into your workflow, you ensure every post is mathematically tuned for maximum reach. Using this tool alongside other Openclaw Skills provides a competitive edge by transforming technical algorithm rules into actionable content improvements.
Tweet Composer Use Cases
Use cases for this skill include:
- Scoring draft tweets before posting to predict reach and engagement.
- Optimizing thread structures to maintain high interest throughout the conversation.
- Reformatting content to avoid algorithmic penalties like link-body placement or hashtag stuffing.
- Analyzing past performance to understand why specific tweets succeeded or failed based on ranking signals.
- Developing a daily posting strategy that respects author diversity scoring and frequency caps.
How Tweet Composer Works
- The skill initializes by referencing the latest algorithmic rules from the provided reference documentation.
- It analyzes the user's draft tweet or thread against a weighted scoring rubric covering replies, media, shareability, and dwell time.
- It identifies negative signals such as excessive hashtags or poor media integration that could trigger a "not interested" prediction.
- The system generates a detailed scorecard including predicted action boosts and specific improvement suggestions.
- Finally, it provides a rewritten, optimized version of the content ready for publication using Openclaw Skills.
Tweet Composer Setup
To set up the Tweet Composer skill, ensure your environment is configured and the reference files are in the correct directory.
# Clone the repository containing the skill
git clone https://github.com/openclaw/skills.git
# Navigate to the tweet-composer directory
cd tweet-composer
# Ensure the algorithm-rules.md is present in references/
ls references/algorithm-rules.md
Then, simply trigger the skill by providing a draft tweet to your AI agent.
Tweet Composer Data Schema & Taxonomy
The skill organizes its analysis and output using a structured taxonomy based on the ranking pipeline:
| Field | Description | Metrics |
|---|---|---|
| Score | 0-100 numerical rating | Weighted average of 7 categories |
| Action Predictions | Likelihood of user engagement | P(reply), P(favorite), P(share), P(dwell) |
| Format Metadata | Structural check | Length, media type, link placement, hashtag count |
| Suggestions | Actionable feedback | List of specific edits for higher reach |
Tweet Composer Advanced Features
- Thread Hook Optimization: Automatically identifies the strongest opening tweet to satisfy the DedupConversationFilter.
- Dwell Time Engineering: Suggestions to increase text length to the 100-200 character sweet spot.
- Negative Signal Mitigation: Proactively flags content that might trigger mute or "not interested" flags.
- Author Diversity Management: Advice on timing posts to avoid the exponential scoring penalty for high-frequency posting.
- Seamless Integration: Works with other Openclaw Skills to automate the entire social media content lifecycle.
SKILL.md
Loading
Related Openclaw Skills
Google Search Skill
A comprehensive Serper.dev integration providing AI agents with rich Google Search results across multiple specialized verticals.
Reddit Research
A powerful, zero-dependency Reddit research agent designed for deep discovery and real-time community monitoring without requiring API keys.
Reddit Research Skill
A powerful, zero-dependency research agent for extracting real-time and historical community insights from Reddit without requiring API keys.
Gemini Image Prompting Skill
A professional-grade tool that converts simple text descriptions into technical, structured JSON prompts for high-fidelity Gemini image generation.
X Research
A professional-grade X/Twitter research agent that provides deep search, profile analysis, and thread extraction through a cost-effective API integration.
Agent Chatroom
A sophisticated social interaction skill for AI agents featuring scrolling danmaku messages, Reddit-style threaded discussions, and community voting.
Featured*
