Search X for Research From Your Terminal1 skill
A command-line interface for querying X/Twitter, monitoring accounts, and pulling threads with cost tracking and caching.
Research Blueprint
Safety checkedSearch X for Research From Your Terminal
rohunvoraMay 2026
A CLI tool wrapping the X API for searching tweets, monitoring accounts, and gathering sourced research without writing API calls.
Search & FilteringThread RetrievalAccount MonitoringCost OptimizationCache Management
Install Command
Run this command to deploy the blueprint to your environment.
What problem does this solve?
Turns Claude Code and other AI agents into X research tools so they can search tweets, pull threads, monitor accounts, and gather recent perspectives without writing curl commands. Teams spend time context-switching to X's web interface instead of collecting research from within their AI workflows.
How does it work?
Once installed with your X API bearer token, you run search commands from your terminal. The tool handles API calls, filtering, sorting, and caching automatically. Commands include search with noise removal and engagement sorting, profile lookups for specific users, thread fetching for full conversations, and watchlist monitoring for key accounts. Output: sourced research with cost transparency for every query.
Included Skill Groups
What's the biggest win?
Research from X flows directly into your AI workflow with automatic cost tracking and result caching, so you never waste API credits on repeat queries.
What should I know technically?
Requires Bun runtime for CLI tooling (native TypeScript execution) and an X API bearer token with prepaid credits. The skill uses file-based caching with a 15-minute default TTL and integrates with Claude Code and OpenClaw via SKILL.md manifest. Token is read from X_BEARER_TOKEN environment variable or ~/.config/env/global.env to avoid inline exposure.
What should I watch out for?
Searches cover only the last 7 days (uses X's recent search endpoint, not full-archive). Bearer token exposure risk when running inside AI agent sessions that log HTTP headers. Requires X API prepaid credits (no free tier), and min_likes/min_retweets operators are filtered post-fetch instead of at query time, which may cost more API credits than expected.
Key Features
Search with Filtering
Query X with engagement sorting (likes, impressions, retweets, recent), time ranges (1h to 7d), and automatic noise removal to surface quality results.
Quick Mode
Fast, cheap lookups that limit results to 1 page, use 1-hour cache TTL, and auto-filter low-engagement noise to reduce API costs.
Watchlist Monitoring
Add accounts to monitor and check recent activity from multiple users at once to track what key voices are saying.
Thread & Profile Retrieval
Fetch full conversation threads by tweet ID or recent tweets from specific users without drilling into each one manually.
Cost Transparency
Every search displays estimated API cost so you see exactly what each query costs and can optimize for cheaper operations.
Tools in this Blueprint
About This Blueprint
- License
- MIT
- Industry
- Technology
- Skills
- 0 workflows, 0 sub-skills, 1 standalone