LLM 與 Agent

Use when creating animations that prompt immediate user action, highlight time-sensitivity, or drive conversions.

Turn the model into a SQLAlchemy-focused Python coding agent for designing models, writing queries, debugging database issues, and integrating SQLAlchemy with Alembic and FastAPI in new or existing Python projects.

Use when building or working with Mastra agents - establishes best practices for agent development, tool creation, and workflow orchestration

HestAI workflow phase execution including D0→D1→D2→D3→B0→B1→B2→B3→B4→B5 sequence, agent assignments per phase, phase transition validation, and deliverable requirements. Use when starting new phase, determining next steps, or validating phase completion.

Use when facing 3+ independent failures that can be investigated without shared state or dependencies - dispatches multiple Claude agents to investigate and fix independent problems concurrently

Transforms stock news into hilarious entertainment scripts using internet memes and slang. Creates cat character dialogue for MeowStreet Wars YouTube Shorts (30-60 seconds). Emphasizes humor, accessibility, and viral potential.

MONAD-grounded cognitive architecture for AI memory as morphemic substrate navigation. Memory is not storage but substrate sampling - accessing the same structure that underlies reality. Implements φ-scaling, GOD operators, toroidal coherence tracking, and the 4.5%/95.5% observable/dark split. Replaces nexus-mind with theoretically grounded architecture.

Claude Code plugin creation guide. Use when creating new plugins, adding commands/skills/hooks/MCP servers to plugins, or setting up marketplace distribution. Triggers on "create plugin", "add command", "add skill", "add hook", "plugin structure".

Maintain skills wiki health - check links, naming, cross-references, and coverage

Setup an autonomous software development cycle, inspired by Anthropic's "Effective harnesses for long-running agents" blog (https://www.anthropic.com/engineering/effective-harnesses-for-long-running-agents). Adapted to fit our tech-stack (TypeScript, NextJS, InstantDB, Trigger.dev, tRPC) and style (kebab-case file-naming).

AI导师框架，通过10+顶级提示工程方法论（C.R.E.A.T.E.、Perfect Prompt Framework、修辞学方法、结构化方法、元提示词理论、迭代优化、个性化定制、逆向思维、系统思维、多模态应用）引导用户创建高效元提示词。当用户需要：(1) 学习提示工程最佳实践，(2) 创建特定任务的元提示词，(3) 优化现有提示词效果，(4) 掌握提示设计思维模式，(5) 建立个性化提示词库，(6) 应用框架解决实际问题，或 (7) 发展AI协作技能时触发。通过'创建元提示词'、'优化提示'、'提示工程指导'、'AI导师'等关键词激活。

Interactive UI theme generation when user needs help choosing colors/fonts.Generates cohesive themes based on brand personality using color harmony theory.Use when user asks for theme/color help OR building UI without design:- "What theme should I use?" "Help me pick colors" "Generate theme"- "What colors work well together?" "Suggest color palette"- User uncertain about design direction, needs aesthetic suggestions- Building UI/landing page with no design specs, needs complete themeKeywords: theme, color palette, colors, fonts, brand personality, color harmonyInteractive workflow: Ask personality → Present options → Generate custom theme.References pre-defined themes in .claude/themes/.Do NOT load for: User has clear aesthetic/colors, Figma/design file provided.Integrates with design-fundamentals: Generates themes following design principles.

Use when the user wants to publish, share, or submit a Claude Code plugin to the marketplace

Implements features and writes production-ready code

Use when launching a new product end-to-end from market research through post-launch monitoring.Orchestrates 15+ specialist agents across 5 phases in a 10-week coordinated workflow includingresearch, development, marketing, sales preparation, launch execution, and ongoing optimization.Employs hierarchical coordination with parallel execution for efficiency and comprehensive coverage.

Generate images using Google Gemini's image generation APIs via Python CLI. Use for "generate image", "create image", "gemini image", "AI image", or when needing AI image generation in Claude Code.

Fast ideation and problem exploration. Triggers on Ukrainian/English phrases like "подумай", "поміркуй", "брейнштормінг", "brainstorm", "think through", "накидай ідеї", "generate ideas". Rapid divergent thinking to uncover insights, problems, and solutions through interactive dialogue.

This skill should be used when creating, modifying, or debugging Claude Code hooks. Triggers on phrases like "create a hook", "add a hook", "make a PreToolUse hook", "automate before tool runs", "block tool execution", "validate tool input", "log tool output", "hook for Write/Edit/Bash", "PostToolUse handler", "session start hook", or when working with files in .claude/hooks/ directory.

AI-powered PR Author Agent that transforms Observability Diff Plans into Pull Requests. Use when: (1) Generating instrumentation code from Scout Agent output, (2) Creating OTel configuration, correlation headers, lineage specs, (3) Scaffolding telemetry validation tests, (4) Creating GitHub/GitLab PRs with observability artifacts. Triggers: "generate PR from diff plan", "create instrumentation PR", "scaffold observability code", "generate OTel config", "create telemetry PR".

Detect when user is working in an un-indexed project and proactively suggest enabling semantic memory. Activates on first code question in new projects to guide users through initial setup.