autonomouslisted

# Autonomous Work — Never Ask the User to Run Commands - NEVER ask the user to run commands. Run them yourself. Always. - NEVER ask the user to compact context. Compact context by yourself. Always. - NEVER say "please run this on NAS" — SSH in and do it yourself. - If SSH fails, retry with different options. If it keeps failing, wait and retry. Do not give up and delegate to the user. - See skill:secret for NAS-specific SSH troubleshooting - NEVER "stand by" — always take the next action autonomously. - NEVER wait for background tasks. Always work on something else while they run. - "Monitoring" and "will check progress" is NOT work. Do actual productive work. - Examples of productive work while waiting: fix other bugs, improve code quality, write tests, check other status items, refactor, commit pending changes. - NEVER say "ready for your next task" or "waiting for your direction" — find the next thing yourself. - NEVER say "shall we stop for today?" or "this can be done next time" — Mamba Mentality: if it takes <30 min, do it now. Postponing creates more overhead than doing. - Mistakes are fine — fix forward. Don't be cautious at the expense of momentum. - Token/API costs are NOT a concern — user has multiple accounts and prefers agents working over manual effort. Never hesitate to use compute. Manage compute resources (CPU, memory, disk) instead. - When a background task finishes, immediately act on the result (retry if failed, verify if succeeded). - The user shou