buckster123

Quick status check and onboarding for ApexAurum - Claude Edition project. Use when starting a new session, checking project status, verifying setup, getting oriented, or when asked about the project structure, what's working, what's pending, or how to continue development.

Launch and manage private LLM endpoints via LocalRouter TUI — local GPU, rented Vast.ai, vLLM multi-GPU clusters, or Together AI managed inference.

Compress a skill file to reduce token count while preserving all technical information. Use whenever the user drops in a SKILL.md, skill file, system prompt, or agent instruction document and asks to "compress", "shrink", "densify", "reduce tokens", "tighten", or "make smaller" — or when they mention a skill is "too long", "bloated", or "costs too much to load". Applies structural compression (collapsing repeated prefixes, telegraphic tables, symbolic notation) to English-prose skills. Target 40-55% token reduction with zero information loss on technical content.