Context Hub serves API docs to AI coding agents via MCP.
Agents call it repeatedly during a session — search, get, search again.
At 1 second per call that lag accumulates. At 56ms it's invisible.
Viet-Anh NGUYEN
@vietanhdev
Machine Learning Engineer
- I rewrote Context Hub in Rust. Search: 1,060ms → 56ms (19x faster) Memory: 122MB → 23MB (5x less) Binary: 22MB + Node.js → 10MB, no deps Drop-in replacement. Same 1,553-doc registry. github.com/nrl-ai/chub 🧵
- A lot of bug fixes and improvements have been made to AnyLearning - the no-code offline AI training tool. We're committed to providing the best tool for our users! Checkout: anylearning.nrl.ai #NocodeAI
- 🎉 Thrilled to share AnyLearning with you all! We've built this with privacy at heart - your data stays on your machine, period. Perfect for ML projects where security matters. 💪 Try it: anylearning.nrl.ai/download #AI #Privacy #BuildInPublic
- Llama Assistant - my offline assistant with Llama 3.2. #LLaMA3
