x-search_skill

Overview

This skill provides X/Twitter search and social listening via the xAI Grok API, packaged as an agent-accessible interface. It helps you find tweets, monitor topics, research people, and surface viral posts. Each search costs 5 credits and returns tweet text, authorship, engagement metrics, and citations.

How this skill works

After a one-time bootstrap to ensure authentication, the skill sends POST requests to the x-search API endpoints with your query, optional date filters, and limits. Responses include content, citations, credits used, and usage metadata. There are dedicated endpoints for general tweet search and for researching a person or handle.

When to use it

• Look up recent tweets and conversations about a product, feature, or competitor
• Track and combine keyword clusters for revenue intelligence or GTM signals
• Find viral posts or high-engagement tweets in an industry vertical
• Prepare quick background and talking points before a meeting with a person
• Set up ad-hoc social listening for crises, launches, or campaign performance

Best practices

• Always run the bootstrap step first to confirm authentication and avoid 401 errors
• Use quoted phrases and OR logic to capture precise language and variations
• Apply days_back for time-bounded searches (e.g., last 7 days) to focus recency
• Filter by engagement (min_faves) and exclude retweets (-is:retweet) to surface originals
• Batch related keyword clusters in separate searches and then merge/rank by engagement
• Monitor credit consumption—each search or person research call consumes 5 credits

Example use cases

• Search: Find the top 20 tweets in the past week mentioning 'AI agents' and sort by likes
• Monitor: Run multiple cluster searches for phrases like 'CRM is dead' and combine results for sentiment analysis
• Research: Pull a concise social profile and recent activity for a meeting with a prospect
• Viral discovery: Query industry keywords with min_faves:50 to find posts worth amplifying
• Revenue intel: Track mentions of 'revenue operations' or 'tech stack' to spot churn signals