发现 Skills

Configures Neovim with the user's opinionated Lua patterns, a namespaced configuration structure, LSP managed via Mason + nvim-lspconfig, and leader-key conventions. Contains user-specific decisions and checklists that Claude cannot know without this skill. Use this skill whenever the user mentions Neovim-related topics — for example: neovim, nvim, init.lua, LSP setup, keymaps, autocommands, Neovim plugins, Mason, TypeScript LSP, slow LSPs, Neovim config, or any Neovim question — even if Claude thinks it can answer from general knowledge.

Use whenever validating a multi-layer app where frontend depends on backend depends on database — Rails+React, Django+Vue, Next.js+Postgres, any monorepo or separate repos that share data. Runs a strict bottom-up protocol: verify DB schema and seed data first, then API CRUD, then frontend rendering, then cross-layer integration flows (create in frontend → appears in DB, insert in DB → appears in frontend, update via API → propagates everywhere, delete cascades). Reach for this when a bug could be in any layer and you need to prove which one, or before shipping a change that touches more than one layer.

Run BEFORE any validation session to catch missing dependencies, dead servers, unseeded databases, and stale builds — the failures that otherwise eat 10–30 minutes of mid-validation debugging. Auto-attempts one fix per failed check (start the DB, install dependencies, run migrations) and re-checks. Produces a CLEAR / BLOCKED / WARN verdict and a structured report at e2e-evidence/preflight-report.md. Reach for it whenever someone says 'check before validating', 'why is the dev server down', 'pre-validation check', 'environment health', before /validate or /forge-execute, or when a previous validation FAILed for environmental reasons (server crashed, DB empty, missing env var).

Use for the design-to-code pipeline via Stitch MCP — generating reference designs from text prompts before implementation, iterating variants, and persisting the design + tokens as the reference that design-validation later compares against. Most useful at the START of a UI feature (explore options visually before writing code) or when you want to lock in a reference for later validation. Pairs with design-validation (compares built UI vs Stitch output) and design-token-audit (checks token compliance). Reach for it on phrases like 'design before code', 'stitch design', 'explore UI options', 'generate UI mockup', 'create design reference', 'design-to-code'.

Rules and checklist for preparing PRs, creating changesets, and releasing packages in the monorepo.

This skill should be used when designing system architecture, API contracts, deployment topologies, or making technology decisions for full-stack applications.

This skill should be activated when reviewing Java code or enforcing coding standards in Spring Boot services. It covers naming conventions, immutability patterns, Optional usage, streams, and exception handling.

CI/CD visual regression setup and pre-commit visual verification checklist. Scoped complement to reality-checker agent — adds: visual regression tool ecosystem (Chromatic, Percy, Applitools, BackstopJS, Playwright Visual), 13-item mandatory verification checklist, and GitHub Actions screenshot diff pipelines. Use when setting up automated visual regression tests in CI/CD or running systematic pre-PR visual verification beyond what reality-checker provides.

Adjusts office temperature.

Helps configure LiteLLM proxy for multi-model routing

This sample demonstrates an Electron debugger hook used to intercept auth responses.

Predicts future balance and simulates purchase impact on finances. Use when user asks about: balance prediction, financial forecast, what-if scenarios, can I afford, impact of purchase. Triggers: 余额预测, 财务预测, 下个月还剩多少, 如果买X, 能不能买, 买得起吗.

Use the az CLI for Azure cloud resource management, Azure DevOps operations, virtual machines (VMs), storage, networking, AKS, and Key Vault with comprehensive authentication and output control.

Semantic and multi-modal search across documents using LanceDB vector embeddings. Use when searching knowledge bases, finding information semantically, ingesting documents for RAG, or performing vector similarity search. Triggers on "search documents", "semantic search", "find in knowledge base", "vector search", "index documents", "LanceDB", or RAG/embedding operations.

Take your AI agent to the next level with full LangWatch integration. Adds tracing, prompt versioning, evaluation experiments, and simulation tests in one go. Use when the user wants comprehensive observability, testing, and prompt management for their agent.

Laravel Fortify headless authentication backend development. Activate when implementing authentication features including login, registration, password reset, email verification, two-factor authentication (2FA/TOTP), profile updates, headless auth, authentication scaffolding, or auth guards in Laravel applications.

Autonomous-only skill (AI-Loop protocol) that merges an approved change request into the base branch, updates task metadata to done, removes review files, and cleans worktrees/branches. Reserved for AI autonomous coding; do not use for human+AI.

Structures REST API routes with Huma v2 and Echo. Use when creating or modifying API endpoints, route handlers, or OpenAPI-backed HTTP APIs in this project.

Performs comprehensive UX testing via Playwright — user journeys, responsive design, accessibility, and interaction bugs. Use when verifying UI quality before a release or after major frontend changes.

Build AI that answers questions about your database. Use when you need text-to-SQL, natural language database queries, a data assistant for non-technical users, AI-powered analytics, plain-English database search, or a chatbot that talks to your database. Covers DSPy pipelines for schema understanding, SQL generation, validation, and result interpretation.

IBM quantum computing framework. Use when targeting IBM Quantum hardware, working with Qiskit Runtime for production workloads, or needing IBM optimization tools. Best for IBM hardware execution, quantum error mitigation, and enterprise quantum computing. For Google hardware use cirq; for gradient-based quantum ML use pennylane; for open quantum system simulations use qutip.

Generate beautiful, self-contained HTML pages that visually explain systems, code changes, plans, and data. Use when the user asks for a diagram, architecture overview, diff review, plan review, project recap, comparison table, or any visual explanation of technical concepts. Also use proactively when you are about to render a complex ASCII table (4+ rows or 3+ columns) — present it as a styled HTML page instead.

Guide for safely discovering and installing skills from external repositories. Use when a user asks for something where a specialized skill likely exists (browser testing, PDF processing, document generation, etc.) and you want to bootstrap your understanding rather than starting from scratch.

Guide for working in parallel with other agents. Use when another agent is already working in the same directory, or when you need to work on multiple features simultaneously. Covers git worktrees as the recommended approach.