
x-post-optimizer
by ElemontCapital
A suite of high-performance AI agent skills derived from the open-source x.AI x-algorithm
SKILL.md
name: x-post-optimizer description: Use this skill when advising on post structure, timing, media selection, and account health to align with the HeavyRanker's scoring weights and anti-spam heuristics. version: 1.0.0 license: MIT
X Post Optimizer
Strategic expertise for maximizing content reach on X. It focuses on the "Engagement Velocity" required to pass from candidate sourcing to the top of the "For You" timeline.
Context
The X algorithm utilizes a WeightedScorer that aggregates multiple probability heads (e.g., $P(\text{Like})$, $P(\text{Reply})$). While the HeavyRanker (Phoenix/MaskNet) predicts the likelihood of an action, the Weights determine the final distribution. Understanding these weights is key to "Algorithm-Native" content creation.
For specific tactical data, refer to:
What it does
- Calculates Expected Value: Estimates the "Score Boost" of different media types (e.g., Video vs. Static Text).
- Optimizes Conversation Depth: Advises on "Author-In-Thread" interactions, which carry massive weight in the
WeightedScorer. - Protects Account Reputation: Identifies "Anti-Signals" (external links, rapid-fire posting) that trigger the
VisibilityLibde-amplification. - Timing Strategy: Leverages knowledge of the "Frequency Deboost" window (3600s) to prevent internal cannibalization of posts.
Guidelines
- The 13.5x Rule: Replies are significantly more valuable than Likes in the modern ranker. Content that invites a meaningful "back-and-forth" creates a feedback loop that the
HeavyRankeroptimizes for. - Author Reply Multiplier: In the code, author engagement on their own thread acts as a "Freshness" and "Conversation" signal, often effectively multiplying the thread's reach by keeping it at the top of the retrieval stack.
- Negative Signal Avoidance: Avoid "Engagement Bait" that leads to "Show Less Often" or "Report" actions. The weight for a "Report" (-369.0) is mathematically impossible to overcome with positive engagement.
- Media Strategy: Native video receives a dedicated weight (
vqv_weight_eligibility), making it the preferred format for "Out-of-Network" discovery.
Example Trigger Prompts
- "/post-check optimize draft: [Insert Post Text]"
- "/post-check why is reach low for this post?"
- "Should I put the link in the main post or first reply?"
- "Why is my second post 50% weaker than the first?"
- "Explain reach differences: video vs text-only"
- "Maximize 'Conversation' weight for 5-tweet thread"
- "Impact of 'Show Less Often' clicks on my account"
Score
Total Score
Based on repository quality metrics
SKILL.mdファイルが含まれている
ライセンスが設定されている
100文字以上の説明がある
GitHub Stars 100以上
1ヶ月以内に更新
10回以上フォークされている
オープンIssueが50未満
プログラミング言語が設定されている
1つ以上のタグが設定されている
Reviews
Reviews coming soon
