What problem does it solve? Gathering real-time opinions, developer discussions, and breaking news from X/Twitter requires writing raw API calls, handling pagination, and filtering noise manually. This Skill wraps the X API into a CLI and an agentic research loop so an AI agent can decompose a question, run targeted searches, follow threads, and synthesize a sourced briefing without boilerplate. ## Core Features & Use Cases - Search CLI with filters: Query recent tweets with engagement sorting, time filters (--since 1h), quick mode, quality filtering, and per-search cost display. - Threads, profiles, and watchlists: Fetch full conversation threads, recent posts from any user, and monitor a stored watchlist of accounts for heartbeat checks. - Caching and cost control: File-based cache (15-minute or 1-hour TTL) avoids repeat API charges, and every search reports its estimated pay-per-use cost. - Use Case: Ask "What are people saying about the new Opus release?" and the agent runs multiple refined searches, follows high-engagement threads, deep-dives linked resources, and saves a markdown research doc with quoted, linked sources. ## Quick Start Ask the agent to search X for what people are saying about a topic, for example: "Do an x research on what developers think about Bun 2.0 and summarize the key opinions with sources."
