🐦
tweet-composer
Score tweets against X's real ranking algorithm
v1.2.0
MIT License
✓ OpenClaw
➜~clawhub install tweet-composer
What it does
X's "For You" feed ranks tweets using 19 engagement signals. This skill scores your drafts against the actual algorithm rules — not generic tips.
How it works
X's "For You" feed is powered by a Grok-based transformer that predicts 19 engagement actions (reply, retweet, favorite, bookmark, dwell time, etc.) for every tweet. Each action has a different weight in the final ranking score. This skill encodes those exact rules into a deterministic scoring engine — your agent runs every draft through the same logic X uses internally, producing a 0–100 score with a full breakdown of what's working and what's not.
Features
📊 Score 0–100 with category breakdown (content, formatting, engagement hooks, timing)
🔮 Predicted action boost: P(reply), P(favorite), P(retweet), P(bookmark)
✍️ Optimized rewrites that maximize reach without changing your voice
🧵 Thread optimization — score each tweet in a thread individually
⏰ Timing and scheduling advice based on engagement patterns
🎯 Hook analysis — first 280 chars get 80% of the ranking weight
Example
output
┌─────────────────────────────────────────┐ │ Tweet Score: 82/100 ⬆️ HIGH │ ├─────────────────────────────────────────┤ │ Content Quality ████████░░ 78/100 │ │ Engagement Hooks █████████░ 91/100 │ │ Formatting ███████░░░ 72/100 │ │ Algorithm Fit ████████░░ 85/100 │ ├─────────────────────────────────────────┤ │ P(reply) 0.34 P(retweet) 0.18 │ │ P(favorite) 0.62 P(bookmark) 0.09 │ ├─────────────────────────────────────────┤ │ 💡 Add a question to boost P(reply) │ │ 💡 First line hook is strong — keep it │ └─────────────────────────────────────────┘