X Daily Report aggregates posts from a curated list of top AI researchers, company accounts, and paper authors — uses an LLM to rank posts by novelty and information value — and outputs a structured daily digest with categories: research papers, hot takes, product launches, hiring, events.
1. **Install and configure:**
npm install -g x-daily-report
x-daily-report configure --accounts ./ai-accounts.txt
2. **Set your digest delivery:**
digest:
format: markdown
delivery:
webhook: your_slack_webhook
channel: "#ai-news"
time: "08:00 UTC"
filters:
include_topics:
exclude_topics:
3. **Run or schedule:**
x-daily-report run # one-shot
x-daily-report serve # runs daily via cron
**Prompt to test it:**
Give me today's AI Twitter digest for topics: LLMs, AI safety. Show me the top 5 most novel posts with engagement stats.
The algorithmic X feed is designed to maximize engagement, not information value. A flame war between two ML researchers will outrank a genuinely important paper announcement. The LLM ranking in this tool explicitly penalises high engagement + low novelty (flame wars, dunking) and rewards direct links to papers, repositories, or blog posts.
| Pros | Cons |
|---|---|
| Curated signal vs. algorithmic noise | X scraping is legally ambiguous |
| LLM ranking is genuinely better than engagement | Free tier has low rate limits |
|---|
| Structured digest saves significant time | Account list requires manual curation |
|---|