For Kraken · AI Infra

Real-time social context, straight into your model.

A 7-minute walkthrough showing how LunarCrush plugs into Claude (or any model with MCP / API) to generate hyper-personalized, asset-specific news in seconds. Same plumbing scales to per-user news rewrites and prediction-market context inside the Kraken stack.

Duration · 6:54 Format · MP4 · H.264 · 1920×540 Captured · 2026-05-13 Download MP4 ↓

What you'll see in the video

  1. The LunarCrush MCP loaded directly into Claude. One-click connector, the same one that's been live in the Claude store since launch.
  2. A natural-language prompt. Joe types: "use LunarCrush to write a personalized news story about Solana for an institutional hedge fund in the 50M–1B AUM range, only last 6 hours, 500 words, render as an HTML dashboard."
  3. Claude calls LunarCrush in real time. The MCP pulls the full payload for Solana, 100K+ mentions, every news link people are sharing, every top creator, sentiment, bigrams and trigrams of co-mentioned terms.
  4. Rendered, formatted output in seconds. An HTML dashboard ready to ship into a product, or be rewritten per-user.
  5. Latency context. A new social post or news event is queryable in 1 second to 1 minute, fast enough for personalized news feeds inside a trading app.