菜单
这个仓库中的 skills
SkillsMP 已收集 CleanExpo/skills-library 中的 150 个 Skill。打开任一 Skill 可查看来源和详情。
CleanExpo/skills-library已展示 40 / 150 个已收集 Skill。
Use when a /goal Stop-hook keeps re-firing the same "Condition unsatisfied" verdict across turns without state changing — detect the unsatisfiable-loop, classify the residual, break it instead of re-emitting reworded status. Fires on repeated judge verdicts…
原文语言：英语
Ingest current session learnings into the Brain-1 wiki at ~/2nd Brain/2nd Brain/Wiki/. Updates existing pages in place, creates new pages for genuinely new topics, appends to log.md. Run at the end of any significant working session.
原文语言：英语
Compete-mode for BUILT artifacts — landing page, component, small app, email template. Type /build-tournament <brief> to have isolated persona contestants each build the full artifact, a blind judge score them, and a revised champion ship. Reserve for…
原文语言：英语
Repurpose one published video into written formats — type /content-cascade <video URL or brand channel> to turn its transcript into a blog post, LinkedIn post, and X post in the brand voice, gated through nexus-copywriter. Drafts only; publish stays…
原文语言：英语
Morning intelligence sweep — type /daily-intel-brief [topics] to pull GitHub trending, X headlines, YouTube momentum (views-vs-subs signal), and web news for the estate's domains into one dated vault brief. Raw signal over analysis.
原文语言：英语
Morning email triage — type /inbox-triage to classify the last 24h of Gmail into leads / urgent / warm / partners / meetings / noise, research real leads, draft (never send) replies, and drop a one-glance brief into the vault.
原文语言：英语
Brand Guardian / Editorial Standards enforcer — quality gate for all non-code output before it reaches a client portal or goes live. Reviews content against the active brand voice for each portfolio business, verifies factual accuracy, enforces the $2B…
原文语言：英语
Use when a task needs AI image or video generation — campaign visuals, hero shots, product or social video, thumbnails, ad creative, localized variants — through the official Artlist MCP (mcp.artlist.io, 100+ models incl. Nano Banana Pro, Seedance 2.0, Kling…
原文语言：英语
Mathematical first principles validation for code quality. Triggers on "optimise", "algorithm", "performance", "complexity", "O(n)", or when reviewing code for efficiency. Enforces Turing (algorithmic), Von Neumann (architecture), Bezier (animation), and…
原文语言：英语
Use when someone asks to build, design, make, create, or generate any deliverable from a short prompt — a deck, web page, report, poster, campaign, brand look, prototype — especially when the prompt is underspecified. Resolves the subject and client persona,…
原文语言：英语
Context health check. Shows token pressure and recommends one action to reduce waste.
原文语言：英语
Production bible for RestoreAssist "Decision Room" episodes — cinematic Remotion compositions at senior production house quality. Use this skill whenever writing, designing, or rendering any Decision Room episode. Covers the full stack: cold open philosophy,…
原文语言：英语
Minimal quick fix. Use for small bugs, type errors, lint issues, or surgical one-file changes. Keeps token usage minimal.
原文语言：英语
Autonomous project orchestration protocol for Next.js full-stack builds. Triggers on "build", "implement", "create feature", "plan", "architecture", or when starting new phases. Enforces phase-locked execution, token economy, and sectional verification gates.
原文语言：英语
Full build pipeline. Use when ready to write code after planning. Injects structured implementation protocol into the current session.
原文语言：英语
Use Chrome browser to interact with Linear visually — read the board, pick up tasks assigned by board members, implement them autonomously. Triggered by "open Linear", "check the board", or "do what's on the Linear board".
原文语言：英语
Turn audit and review findings into Linear issues in the right project, and define how builder agents pull work from Linear and keep building until the backlog is empty. Use when the user asks to send findings to Linear, sync tasks, or kick off autonomous…
原文语言：英语
Task transition. Use after completing a task to determine and prepare for the next one. Reads PROGRESS.md and Linear.
原文语言：英语
Use when optimising images in a Next.js / React codebase for SEO or accessibility — adding priority to above-fold next/image components, applying aria-hidden to decorative SVGs in JSX, or converting logo assets to WebP. Triggers on "priority prop",…
原文语言：英语
Use when a Unite-Group Nexus app (Unite-Hub / Command Centre, RestoreAssist, Synthex, etc.) shows connector/env/auth symptoms — "No API key found in request", a login/OAuth redirect loop, a dev-server restart loop ("Found a change in next.config.js"),…
原文语言：英语
Run a strategic output through the 9-persona CEO Board before it distributes. Invoke whenever Pi-CEO is about to push content, create external-facing PRs on Urgent/High issues, or act on a strategic recommendation.
原文语言：英语
Architecture planning with deep reasoning. Use for "plan", "design", "architect", "how should we build". Read-only, structured plan output. Runs in a forked context with Opus.
原文语言：英语
Use when checking CI status on a GitHub PR and deciding whether to merge — distinguishes blocking failures (GitHub Actions quality gates) from ignorable failures (Vercel preview deployments, sandbox Vercel builds). Triggers on "CI failing", "Vercel failing on…
原文语言：英语
Use when working on any App Store or Google Play submission task for RestoreAssist — enrollment, certificates, store listings, testing tracks, or release submission
原文语言：英语
Use Chrome browser to complete YouTube Studio channel rebrand for RestoreAssist (RA-238)
原文语言：英语
Use when creating, editing, or fixing scheduled tasks on claude.ai/code/scheduled for RestoreAssist cron endpoints
原文语言：英语
Snapshot recovery. Use at session start, after compaction, or when resuming previous work. Restores state from PROGRESS.md and snapshots.
原文语言：英语
Isolated code review. Use for "review this", "check this code", "audit changes". Read-only, prioritised findings. Runs in a forked context.
原文语言：英语
Design system enforcement for Scientific Luxury tier UI. Triggers on "design", "UI", "component", "styling", "animation", or when creating React components. Enforces OLED black backgrounds, spectral colours, single-pixel borders, physics-based animations, and…
原文语言：英语
Scope a task by pulling the latest note from the user's Obsidian "2nd brain" vault, gather related context from the vault, run a client-friendly multiple-choice dialog with recommendations, and produce a plan/spec (no production code). Use when the user says…
原文语言：英语
Use when the user asks to browse, click, navigate, fill a form, screenshot, or read content from a live website via the claude-in-chrome MCP tools — or whenever a task needs to drive an actual Chrome tab rather than fetch static HTML.
原文语言：英语
Use before opening any PR, pushing to a shared branch, or merging in this estate — where automation force-readies draft PRs and squash-merges them on green within minutes, so opening a PR is authorising its merge. Fires when about to run gh pr create / gh pr…
原文语言：英语
Use when a creative brief has multiple plausible directions and picking wrong is costly — landing-page hero copy, taglines, headlines, hooks, video concepts, design directions. Generate several diverging variants, judge them independently, synthesise a…
原文语言：英语
Multi-CLI agent farm orchestrator. Manages Claude Code (3x Max Plan accounts) and OpenAI Codex CLI (1x Max/Plus account) as tmux-based background workers. Receives routing requests from Pi-Dev-Ops /nexus, dispatches tasks, monitors health, and returns…
原文语言：英语
Use when starting ANY non-trivial development work — a feature, integration, refactor, or bugfix bigger than one file — BEFORE writing code. Runs the spec-driven pipeline (constitution → specify → clarify → judge → plan → tasks → analyze → implement →…
原文语言：英语
The Tao is the autonomous execution engine that ties self-direction with governance. It orchestrates /nexus (routing), /tao-loop (iterate-generate-deploy), /tao-judge (iterative-evaluator-calibrator), and /session-handoff (continuity/escrow) into a single…
原文语言：英语
Weekly cross-repo self-improvement loop. Every Monday 02:00 AEST it applies the 8-Claude-Loops method (INGEST / BUILD / COMPOUND + North Star) across every repo in .harness/projects.json, opening review PRs so all projects compound over time. Cost-optimised…
原文语言：英语
Senior-review-board audit of the 2nd Brain vault — challenge shortlisted ideas, assign dispositions, route survivors to storm/spm/skill-authoring. Writes only inside its _system/wiki-growth/ sandbox.
原文语言：英语
Use when a task needs to control a native macOS desktop app that has no dedicated MCP or web UI — DaVinci Resolve, Pro Tools/Logic, Finder, Maps, System Settings, or any other native app — via screenshot-and-click automation, or when the user asks to see or…
原文语言：英语
Executes the production merge after Phill's explicit approval. Verifies CI, confirms scope, merges to main, triggers deployment, closes mandate. Called when Phill says "approve PR-[number]" to Margot. Model: claude-sonnet-5 (production operation — senior tier)
原文语言：英语