x-research-skill
github.com/rohunvora/x-research-skill- Category
- AI Agents
- Rank
- No. 729Tools index
Previous survey · No. 720 ·
- Pricing
- Open Source
- Platform
- cli
- Type
- TOOL
- Builder
- @rohunvora
- GitHub
- 1.2k stars
- Date
About
A command-line tool that wraps the X/Twitter API for AI agents like Claude and OpenClaw to perform research tasks. Enables searching tweets, following threads, monitoring accounts, and generating sourced briefings with engagement sorting and cost transparency.
What it does
It turns a research question into a repeatable terminal workflow. Searches can target time windows and authors, remove common noise, rank results by engagement, follow conversations, and produce JSON, compact text, or Markdown. A watchlist supports recurring checks, while local caching reduces duplicate requests.
Why it's ranked here
The case is strong for focused, recent social research. It combines useful query controls with source links, readable exports, cache behavior, and visible cost estimates. The included agent instructions also teach query refinement and thematic synthesis. Its narrow archive window and paid API requirement keep it from replacing broader research tools.
What's good
Quick mode imposes sensible budget discipline through one-page retrieval, fewer displayed results, reply and repost filtering, and a longer cache window. Engagement filters operate after retrieval because the API lacks those search operators. Results retain author, timing, metrics, extracted links, and direct post URLs, which supports traceable summaries.
Tradeoffs
Search reaches only the recent seven-day endpoint, despite documenting full-archive availability. Engagement thresholds reduce displayed results after posts have already been fetched, so they do not reduce read charges. Account checks can become expensive because each profile needs a user lookup plus post retrieval. It also requires Bun, prepaid X API access, and careful bearer-token handling around agent logs.
How to use it well
Use it for developers, analysts, or agents needing a fast pulse on recent technical discussion, product feedback, expert commentary, or selected accounts. Start with quick mode, refine noisy queries, then follow valuable conversations and inspect linked sources before synthesizing themes. Keep it read-only and pair it with web research for older history, source verification, and material outside X.
Technical notes+
x-search.ts is the Bun CLI entry point and coordinates argument parsing, watchlists, cache access, filtering, formatting, saving, and estimated costs. lib/api.ts calls X API v2 with bearer authentication, normalizes posts into a typed Tweet shape, paginates recent search, and provides thread, profile, and single-post retrieval. lib/cache.ts stores JSON entries under data/cache/ using short MD5-derived keys and TTL expiry. lib/format.ts emits Telegram-oriented text and Markdown research documents. SKILL.md supplies the agentic research loop, while references/x-api.md documents endpoints and operators. Saving assumes ~/clawd/drafts/ already exists because the shown code does not create it.
Observed
- License
- MIT
- Primary language
- TypeScript
- Install surface
- Git clone into a Claude Code or OpenClaw skills directory, then run with Bun
- Interfaces
- Command-line interface plus agent skill instructions
- API scope
- Read-only X API v2 access using bearer-token authentication
- Output formats
- JSON, Markdown, and compact Telegram-oriented text
- Storage
- File-based JSON cache and watchlist
Read from README.md, lib/api.ts, lib/cache.ts, lib/format.ts, x-search.ts, SKILL.md, CHANGELOG.md, references/x-api.md, data/watchlist.example.json.
What it can do
Search tweets by keywords or phrases
Search query terms → Filtered tweet results with engagement data
Pull complete Twitter conversation threads
Tweet URL or thread identifier → Full thread conversation chain
Monitor specific Twitter accounts for activity
Twitter account usernames or handles → Account activity updates and new tweets
Generate sourced research briefings from Twitter data
Research topic or query parameters → Structured briefing document with tweet sources
Sort tweets and threads by engagement metrics
Collection of tweets or search results → Ranked list ordered by likes, retweets, and replies
Track API usage costs and spending
API requests and operations performed → Cost breakdown and usage transparency reports
Tags
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.