LucioLiu
UserOne-command assembly for a portable AI agent team you own - Nuwa, templates, team rules and memory, wired with preview/apply/verify/revert transactions.
Categories
Indexed Skills (7)
dont-reinvent
Teach an agent to acquire capabilities on its own — find existing skills before building, install to the right layer, adapt freely, sediment good methods back. Use when the agent faces a task needing a capability it doesn't have, or says "I can't do X" / 需要新能力/没有这个工具/想装 skill 时触发。
track-record
A calibration loop for material, genuinely uncertain AI judgments whose outcomes can be observed later - log the bet with a numeric probability and an observable trigger, settle against real evidence, and distill recurring biases into rules. Do not trigger for deterministic/trivial claims or claims with no observable settlement event. Use when logging or settling a material prediction, running a calibration review, or when track record, prediction ledger, settle bets, or learning from outcomes comes up. 中文触发：材料性判断登记、复盘、回填、结算、判断校准、校准回顾、赌局台账。
user-md
Keep your AI's understanding of you in a Markdown file you own - a five-section USER.md with a hand-written lane and an append-only agent-curated lane (dated, sourced, confidence-tagged), machine-local facts split into a gitignored USER.local.md. No silent rewrites; moves when your folder moves. Use when setting up a USER.md or user profile for an agent, recording user preferences or working style, deciding which file a fact about the user belongs in, curating agent-observed habits with confidence tags, onboarding a new agent to an existing user, or carrying personalization across machines and harnesses. 中文触发：用户画像、用户偏好沉淀、记住我的习惯、USER.md 建档养护。
nomad
Turn any folder into a portable, multi-harness AI team workspace — build a new team from scratch, migrate an existing scattered setup into a clean skeleton, or mount an already-built team onto a new AI harness (Claude Code, Codex/ChatGPT, OpenCode, Copilot, Cursor). Trigger on requests like "make my AI team portable", "give my agents persistent memory across tools", "set up a multi-agent team folder", "migrate my prompts/config into a team structure", "connect this team to another AI tool", "sync my agent across tools", "AI team memory portability", or "switch AI tools without losing my agent's memory".
dont-reinvent
Teach an agent to acquire capabilities on its own — find existing skills before building, install to the right layer, adapt freely, sediment good methods back. Use when the agent faces a task needing a capability it doesn't have, or says "I can't do X" / 需要新能力/没有这个工具/想装 skill 时触发。
user-md
Keep your AI's understanding of you in a Markdown file you own - a five-section USER.md with a hand-written lane and an append-only agent-curated lane (dated, sourced, confidence-tagged), machine-local facts split into a gitignored USER.local.md. No silent rewrites; moves when your folder moves. Use when setting up a USER.md or user profile for an agent, recording user preferences or working style, deciding which file a fact about the user belongs in, curating agent-observed habits with confidence tags, onboarding a new agent to an existing user, or carrying personalization across machines and harnesses. 中文触发：用户画像、用户偏好沉淀、记住我的习惯、USER.md 建档养护。
track-record
A calibration loop for material, genuinely uncertain AI judgments whose outcomes can be observed later - log the bet with a numeric probability and an observable trigger, settle against real evidence, and distill recurring biases into rules. Do not trigger for deterministic/trivial claims or claims with no observable settlement event. Use when logging or settling a material prediction, running a calibration review, or when track record, prediction ledger, settle bets, or learning from outcomes comes up. 中文触发：材料性判断登记、复盘、回填、结算、判断校准、校准回顾、赌局台账。
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.