luxueliu-intel-scout

Luxueliu-intel-scout turns daily information overload into a quiet, hands-free pipeline. It reads public RSS and Atom feeds on two dedicated tracks. The AI track focuses on large models, embodied intelligence, and compute, hard-excluding noise like consumer electronics or game trailers. The neuroscience track follows arXiv, Nature Neuroscience, and Neuron. After keyword filtering and deduplication, a local OpenAI-compatible LLM gateway compresses the candidates into a quick-brief plus detail digest, written to a local markdown file. No private data ever leaves your machine.

Built-in PowerShell scripts register Windows Task Scheduler entries with a primary run and a two-hour catch-up run, so a fresh brief is waiting when you open your laptop. Optional push notifications via ServerChan are silently skipped when no key is set. The codebase is pure Python 3.12 standard library with zero third-party dependencies.

Ideal for AI engineers and researchers, neuroscience and consciousness students, and anyone who wants a three-minute daily scan of two focused feeds without drowning in the information firehose. Secrets are read only from environment variables or a local secrets.env file that is gitignored by default.