Skills
98,873 packaged capabilities and playbooks for agents. Install any of them with one command.
Development
Complete guide to Claude Code plugin development, architecture, directory layout, components, and marketplace publishing. PROACTIVELY activate for: (1) creating a plugin from scratch, (2) building or scaffolding a plugin, (3) writing plugin.json, (4) adding commands/agents/skills/hooks/MCP servers, (5) packaging code as a plugin, (6) publishing to a marketplace, (7) validating plugin structure, (8) marketplace.json registration, (9) cross-platform compatibility, (10) version and metadata sync. Provides: plugin schema, layouts, workflows, and publishing steps.
Provides comprehensive guidance for React Hooks including useState, useEffect, useContext, useReducer, useMemo, useCallback, custom hooks, and hook patterns. Use when the user asks about React Hooks, needs to use hooks in functional components, create custom hooks, or optimize hook performance.
React performance optimization patterns, including memoization, code splitting, bundle size reduction, re-render elimination, and profiling. Covers React Compiler automatic optimization, manual memo/useMemo/useCallback targeting, React.lazy with Suspense, barrel file avoidance, content-visibility for large lists, startTransition for non-urgent updates, and React DevTools profiling. Use when optimizing React app performance, reducing bundle size, eliminating unnecessary re-renders, debugging slow components, code splitting, or profiling rendering bottlenecks. Use for performance audit, bundle analysis, re-render diagnosis, lazy loading, virtualization.
20 years of expertise as a senior/lead React developer with the TanStack ecosystem (Query, Router, Table, Form, Start). Use this skill when: (1) Building React applications with TanStack libraries, (2) Reviewing/refactoring React code for clean code, (3) Debugging React/TanStack issues, (4) Setting up a maintainable project structure, (5) Optimizing React app performance, (6) Choosing the right library for a specific use case, (7) Preventing common bugs and memory leaks, (8) Implementing KISS and less-is-more best practices. Trigger keywords: React, TanStack, React Query, TanStack Router, TanStack Table, TanStack Form, TanStack Start, Vinxi, clean code, refactor, performance, debugging.
Turn broad requirements, large stories, epics, features, initiatives, or backlog items into small end-to-end child stories without turning them into technical component tasks. Use when refining a backlog, decomposing epics, planning an MVP or walking skeleton, looking for vertical slices, reducing story size, applying SPIDR/Hamburger/capability slicing, avoiding scatter-gather/component stories, or deciding the first valuable story before implementation planning
Load this skill whenever the project contains SVG graphics — inline SVGs, external SVG files, SVG icons, SVG illustrations, or SVG-based data visualizations. Under no circumstances use SVG without proper accessible titles, descriptions, and ARIA roles where required. Absolutely always add <title> and <desc> to meaningful SVGs and set role="img" with aria-labelledby pointing to those elements. Also apply this skill whenever the project accepts untrusted or user-uploaded SVG content, which requires sanitization before rendering.
Implement Syncfusion WPF PropertyGrid for browsing and editing object properties with custom editors and category editors. Use this when building property inspectors, property panels similar to Visual Studio's Properties window, or dynamically inspecting and editing object properties in WPF. Covers custom value editors, category editors, collection editing, and design-time property editor customization.
Comprehensive guide for implementing Syncfusion WPF Sparkline (SfSparkline) controls in Windows Presentation Foundation applications. Use this when working with sparklines, mini charts, or trend visualization. This skill covers sparkline types (line, column, area, WinLoss), markers, track ball, range bands, axis controls, and segment customization for compact data visualization in WPF applications.
Implement Syncfusion WPF SfStepProgressBar for multi-step process visualization. Use this when building wizard steps, order tracking, stepper indicators, checkout flows, registration wizards, or onboarding step sequences in WPF. Covers StepViewItem, SelectedIndex, SelectedItemStatus, step marker customization, connector lines, data binding, and template selectors based on step status.
Use when a Telegram job needs a user account, not a bot token: history a bot cannot see, acting as a person, bulk export over MTProto, and the ban risk. Not for bots (telegram-bots) or Mini Apps (telegram-miniapps). Opt out: no telegram, без телеграма. Triggers - "telethon", "userbot", "юзербот".
Use when implementing mouse picking, hover detection, object selection, or line-of-sight checks in Three.js. Prevents the common mistake of raycasting every mousemove frame, not using layers for filtering, or missing recursive flag. Covers Raycaster, setFromCamera, intersection format, InstancedMesh picking, layers, performance. Keywords: Raycaster, mouse picking, intersectObjects, click, hover, setFromCamera, intersection, instanceId, object selection, click on 3D object, detect mouse over, pick element.
Use when adding sound to a Three.js scene: background music, 3D spatial audio, or audio visualization. Prevents the common mistake of ignoring browser autoplay policy, not attaching AudioListener to camera, or wrong distance model. Covers AudioListener, Audio, PositionalAudio, AudioAnalyser. Keywords: audio, sound, AudioListener, PositionalAudio, 3D audio, spatial audio, music, AudioAnalyser, Web Audio API, play sound, add music, sound in 3D scene.
Use when adding physics simulation to a Three.js scene: rigid bodies, collisions, constraints, or raycasting. Prevents the common mistake of wrong Vec3/Vector3 conversion, missing world.step(), or choosing the wrong engine. Covers cannon-es and Rapier with Three.js sync patterns, performance comparison. Keywords: physics, cannon-es, Rapier, rigid body, collider, collision, gravity, simulation, WASM, constraint, joint, objects fall, collision detection, bouncing, realistic movement.
Procedural geometry compilation for Three.js WebGPU/TSL. Use when a task needs an indexed mesh writer, contour-derived terrain or terraces, profile or branch sweeps, or a BatchedMesh, InstancedMesh, or dynamic-update decision.
Manage SendGrid email suppressions: bounces, blocks, spam reports, invalid emails, global unsubscribes, and ASM suppression groups. Covers when and how to remove suppressions, reputation impact, and category-based unsubscribe management. Use when debugging SendGrid delivery issues or building unsubscribe flows. Requires a SendGrid API key (SG.-prefix) — not applicable to the Twilio Email API (comms.twilio.com).
Guide for adding a new RPC to personhog-replica and personhog-router. Covers eligibility checks, proto definition, code generation for Python and Node.js clients, Rust implementation (storage trait, postgres queries, service handler, router wiring), and index compatibility validation. Use when adding a new gRPC endpoint to personhog, migrating a Django ORM query to personhog, or extending the personhog service API.
Advanced T-SQL patterns and techniques for SQL Server and Azure SQL. Proactively activate for: (1) writing CTEs and recursive queries, (2) the APPLY operator (CROSS APPLY, OUTER APPLY), (3) the MERGE statement and OUTPUT clause, (4) temporal tables (system-versioned), (5) In-Memory OLTP / memory-optimized tables, (6) advanced grouping (ROLLUP, CUBE, GROUPING SETS), (7) PIVOT and UNPIVOT, (8) ROW_NUMBER/RANK/DENSE_RANK with PARTITION BY, (9) hierarchies via recursive CTE or HierarchyId, (10) JSON_ARRAYAGG and STRING_AGG patterns. Provides a pattern catalog with examples for each technique, performance tradeoffs, and copy-pasteable templates for common advanced query shapes
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.
Generate a CawPlan status report over a date range: for a single product (progress, risk analysis, priority recommendations, summaries), for a Team (CawPlan product line), or for a named member — ticket-change-based completion, in the last two cases. Also lists product-wide open tickets by a requested label-backed category and optional title keyword, including backlog and cross-version tickets. Use when: the user asks for a product status report, progress report, risk summary, release readiness, priority recommendations, the open/unclosed tickets in a product's current or named version, or a product-wide filtered list such as "remaining Collage bugs"; asks how a Team/product line is doing over a date range; asks which product/Team/version tickets were reopened over a period; or asks how a specific member's task completion looks over a date range (not their own — use `cawplan-my-work` for "my tasks"). NOT for: raw activity feed, user activity, ticket creation, metrics dashboards, or critical issue lists.
Use when writing, generating, or optimizing SQL for CockroachDB, designing CockroachDB schemas, or when the user asks about CockroachDB-specific SQL patterns, type mappings, and distributed database best practices. Also use when encountering CockroachDB anti-patterns like missing primary keys, sequential ID hotspots, or incorrect type usage.
Open a GitHub pull request that matches what the branch ACTUALLY changed — validated against the real base, not the branch name or an optimistic summary. Reads the full diff against the true base branch, blocks on protected-branch/empty-branch/conflict-marker states, invokes `commit` first if the tree is dirty, classifies type/scope/breaking-changes from the diff, runs the narrowest relevant checks, pushes the branch, then creates the PR with `gh` and verifies its number, URL, and state. Never fabricates passing test status or an absent breaking change, and never creates the PR before the branch is on remote; falls back to a prepared title/body/compare URL if `gh` is missing. Use only when the user explicitly asks to open or create a PR.
Debugs Windows and Linux/macOS applications (native, .NET/CLR, mixed-mode) with WinDbg MCP (crash dumps, !analyze, !syncblk, !dlk, !runaway, !dumpheap, !gcroot, BSOD), dotnet-dump, lldb with SOS, createdump, and container diagnostics (Docker, Kubernetes). Covers hang/deadlock diagnosis, high CPU triage, memory leak investigation, kernel debugging, and dotnet-monitor for production. Spans 17 topic areas. Do not use for routine .NET SDK profiling, benchmark design, or CI test debugging.
Browse Feishu OpenAPI methods locally and pass through calls to any API. Use the three-part `service.resource.method` schema to look up paths, parameters, and scopes without going online; call any endpoint directly with `feishu-cli api <method> <path>` (v1.29+, covering 2500+ unwrapped APIs). Use when users ask whether Feishu has an XX API, what parameters an X API takes, what scope an X method requires, to browse OpenAPI methods, see how the SDK makes a call, or call an unwrapped Feishu API. Not applicable: checking the latest online schema (use OpenAPI Explorer; local metadata may lag behind the Feishu Open Platform).
Fix Content Security Policy (CSP) compliance issues that UI5 linter reports but cannot auto-fix. Use this skill when linter outputs: - `csp-unsafe-inline-script` with message "Use of unsafe inline script" Trigger on: inline JavaScript in HTML files, script tags without src attribute, onclick/onload handlers in HTML. Also use when user mentions 'security headers', 'inline script warning', 'CSP policy', 'unsafe-inline', or 'Content-Security-Policy'. Provides guidance on moving inline scripts to external files for CSP compliance.
Fix runtime-breaking patterns in JS files that the UI5 linter does NOT report. Use this skill when: - Tests fail after modernization with "ReferenceError", "TypeError: Cannot read properties of undefined", or "X is not defined" - The linter shows zero remaining errors but tests still crash at runtime - Phase 3, Step 3.2 of the modernize-ui5-app workflow (runs after the parallel globals batch, before fix-cyclic-deps) - User mentions "runtime errors after modernization", "test failures after modernization", "global namespace still used" - After running fix-js-globals and fix-cyclic-deps, but tests STILL fail - User sees "X is not a constructor" or module loading errors in test output The linter only checks sap.* globals in JS. App-specific namespace patterns (e.g., com.example.app.utils.Module) are invisible to it. This skill detects and fixes those across ALL JS files — app source AND test. NOTE: App-namespace globals in XML files are handled by `fix-xml-globals` — NOT by this skill.
Fix partially deprecated API usage that UI5 linter reports but cannot auto-fix. Use this skill when linter outputs: - `no-deprecated-api` for Parameters.get, JSONModel.loadData, Mobile.init, ODataModel.v2.createEntry, View.create, Fragment.load, Router constructor - `unsupported-api-usage` for string formatter in JS bindings Trigger on: Parameters.get without object param, JSONModel.loadData with bAsync=false, Mobile.init with homeIcon, createEntry with batchGroupId, View.create/Fragment.load with non-XML type, Router without async:true, formatter as string in JS. Provides specific fix patterns for each partially deprecated API variant.
Conventional Commits format for git commits and PR/MR titles. Type prefixes, scope rules, breaking change syntax, and commit message structure. Use when committing changes, writing commit messages, creating PR/MR titles, or formatting squash merge messages. Not for PR workflows (git-pr), CI/CD status (git-ci), or git branch management
GitHub AI-powered security and automation features (2025-2026). Proactively activate for: (1) GitHub Copilot in PRs (Copilot code review, Copilot autofix), (2) GitHub Advanced Security (CodeQL, secret scanning, dependency review), (3) Copilot Workspace, (4) GitHub Models (gh models eval, prompt evaluations), (5) GitHub Copilot CLI 2.x, (6) AI-assisted issue triage, (7) Copilot extensions and skillsets, (8) Copilot Enterprise context (knowledge bases), (9) automated dependency updates (Dependabot AI grouping), (10) Spark for app generation. Provides a feature catalog, Copilot CLI command reference, gh models eval examples, GHAS configuration, and Copilot Workspace usage
AI movement decision router for chase, patrol, crowd, and bake choices on top of NavigationAgent/Server. Use when deciding node agent vs RID server, bake vs obstacle, layer masks, or retarget policy — not for engine navmesh recipes. Keywords: AI navigation, chase retarget, patrol, crowd RVO, bake vs obstacle, NavigationAgent decision tree.
Drupal security and coding rules from Ivan Grynenko's cursorrules, organised by OWASP Top 10 (2021) category: authentication, access control, injection (SQL, XSS), cryptographic failures, security misconfiguration, vulnerable dependencies, SSRF, insecure design, integrity, and logging, plus database standards and file permissions. Use when writing or reviewing Drupal code for security, handling user input, building queries, handling file uploads, adding permission or access checks, making outbound HTTP requests, or preparing a security review.
Design and build native-feeling macOS application UIs. Use this skill whenever the user asks to create a desktop app, macOS app, Mac-style interface, Apple-style UI, system utility, or anything that should look and feel like a native Mac application. Also trigger when users mention “native feel,” “desktop app design,” “Apple design patterns,” “sidebar layout,” “traffic lights,” or want to build tools/utilities that feel like they belong on macOS. This skill covers layout, composition, interaction patterns, animations, light/dark mode, and all the subtle details that make an app feel like Apple built it
Microsoft .NET Microservices Architecture Guide reference. Proactively activate for: (1) reading or applying patterns from Microsoft .NET Microservices Architecture for Containerized .NET Applications, (2) eShopOnContainers reference architecture, (3) DDD and bounded contexts, (4) integration events (RabbitMQ, Azure Service Bus), (5) API gateway patterns (Ocelot, YARP), (6) resilience patterns (Polly), (7) data patterns (CQRS, event sourcing), (8) testing strategies in microservices, (9) deployment to Kubernetes/AKS/Container Apps. Provides a chapter-by-chapter summary, code patterns from eShopOnContainers, decision matrices for sync vs async integration, and links to specific guide sections
Migrate Python LangChain Deep Agents applications to Pydantic AI and Pydantic AI Harness. Use when the source uses the upstream `deepagents` package or demonstrably reproduces its middleware, backends, skills, memory, subagents, or sandbox contracts. Use `migrating-langchain-to-pydantic-ai` for plain LangChain, LangGraph, or LCEL migrations and `pydantic-ai-harness` for greenfield Harness usage.
Enforces reproducible derived content — output every device must compute identically from a key, with no server. The key is injected, never read from a clock; a calendar-day key is a civil date, not an instant, and one day definition serves every comparison; the seed is a specified hash plus a mix step feeding a PRNG you own, never `dart:math`'s `Random`; entropy has exactly one source, so no `DateTime.now()`, no ambient `Random()`, no ordering by `identityHashCode`; content is regenerated from (key, generatorVersion) rather than stored; a shipped generator is frozen and an improvement is a versioned cutover, never an in-place edit that rewrites history; and it is pinned by a committed golden-vector table of fingerprints from an independent oracle, regenerated only by a reviewed command CI verifies but never blesses. Use when generating content from a seed or date, writing or regenerating golden vectors, versioning a generator, or debugging output that differs between devices or runs.
This skill should be used when the user asks about “Unity performance”, “optimization”, “GC allocation”, “object pooling”, “caching”, “Update loop optimization”, “memory management”, “profiling”, “framerate”, “garbage collection”, or needs guidance on performance best practices for Unity games
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.