Skills
101,174 packaged capabilities and playbooks for agents. Install any of them with one command.
All
Plan yield farming and CAKE staking on PancakeSwap. Use when user says "farm on pancakeswap", "stake CAKE", "unstake CAKE", "stake LP", "unstake LP", "yield farming", "syrup pool", "pancakeswap farm", "earn CAKE", "farm APR", "harvest rewards", "deposit LP", "withdraw LP", or describes wanting to stake, unstake, or earn yield on PancakeSwap. Do NOT use this skill when the user has a token pair and asks how to earn yield or best profit with those tokens (prefer liquidity-planner instead).
Build Python APIs with FastAPI, Pydantic v2, and SQLAlchemy 2.0 async. Covers project structure, JWT auth, validation, and database integration with uv package manager. Prevents 7 documented errors. Use when: creating Python APIs, implementing JWT auth, or troubleshooting 422 validation, CORS, async blocking, form data, background tasks, or OpenAPI schema errors.
Force-links a development or target environment to a Power Platform Pipelines host, overriding any existing association with a previous host. Use when creating a deploymentenvironments record fails with "this environment is already associated with another pipelines host", or when intentionally migrating an environment from one host to another (e.g., Platform Host → Custom Host, or between two Custom Hosts). Calls the documented `ManageEnvironmentStamp` Dataverse action (the API behind the "Force Link" button in the Deployment Pipeline Configuration app). DESTRUCTIVE to the previous host: makers lose access to any pipelines in that host that used this environment. Reversible by running Force Link from the previous host. Use when asked to: "force link environment", "force-link to new host", "switch pipelines host", "environment already associated with another host", "take over pipelines association", "relink environment to host".
Tour — Claude tự diễn cho user xem harness chặn mình theo thời gian thực trên project thật. Hai độ dài — "short" = 3 cảnh (R1·R2·R3), "full" = 10 (R1–R10). Tự phát hiện PoC vendor-neutral (cách CHÍNH) hay harness production. Tự dọn sạch demo. Trigger: "harness tour", "xem harness làm gì", "/harness-tour", "test short", "test full".
Provides role-specific expertise for 26+ IT company positions. Use when responding from a specific role's perspective, providing role-based advice, or when specialized knowledge is needed for engineering, product, design, data, marketing, sales, operations, HR, legal, finance, leadership, or academic functions.
Apply Andrej Karpathy AI methodology and principles from his 2023-2026 insights. Use this skill when the user wants to apply Karpathy-style thinking, needs guidance on agentic engineering, LLM knowledge bases, minimalist coding, understanding-first principles, or any of the 14 core methodologies distilled from his work. Also triggers on: "karpathy method", "karpathy approach", "karpathy way", "AK style", "like karpathy does", "karpathy principles", or when user asks how Karpathy would approach any programming or AI problem.
Use 卖大律 to detect potential TRO (temporary restraining order) and intellectual property (trademark/patent/copyright) infringement risks for products. Provide a main product image (image URL or Base64 data URI supported), with optional reference images, reference text, and IP keywords. Returns an overall risk level, high-risk infringement items and low-risk IP list (including TRO plaintiffs, filing dates, court case numbers, and case counts), a numeric risk score from 0–10, and an AI-generated legal assessment report. Trigger this skill when users mention TRO detection, TRO risk, TRO infringement, IP infringement detection, trademark infringement, patent infringement, copyright infringement, IP risk detection, product compliance checks, 卖大律, product TRO detection, IP infringement risk, or trademark/patent/copyright infringement checks. Also trigger it when a user provides a product image and wants to assess infringement risk related to trademarks, patents, copyright, or TRO, even without explicitly mentioning 卖大律 or TRO.
Comprehensive LLM security assessment against OWASP Top 10 for LLM Applications 2025. Use when reviewing LLM-integrated applications, RAG pipelines, chatbots, AI agents, or GenAI features. Covers prompt injection, data poisoning, supply chain, excessive agency, and more with real-world attack scenarios and testing methodologies.
Marketing strategy expertise for brand strategy, market research, customer segmentation, digital marketing, content strategy, marketing analytics, campaign management, and PR/communications. Use when planning campaigns, analyzing brand health, building personas, or optimizing marketing performance.
Integration guide for Morph's WarpGrep (fast agentic code search) and Fast Apply (10,500 tok/s code editing). Use when building coding agents that need fast, accurate code search or need to apply AI-generated edits to code efficiently. Particularly useful for large codebases, deep logic queries, bug tracing, and code path analysis
Comprehensive red team operations methodology covering full engagement lifecycle from planning through reporting. Addresses engagement scoping and rules of engagement negotiation, multi-tier C2 infrastructure design with redirectors and domain fronting, malleable traffic profiles and beacon tradecraft, OPSEC discipline including attribution avoidance and indicator management, EDR and AMSI evasion techniques using direct syscalls and unhooking, data collection with chain-of-custody controls, and structured reporting with purple team debrief workflows. Covers assumed-breach, external-to-internal, insider threat, and hybrid physical-cyber engagement scenarios with MITRE ATT&CK mapping throughout. Targets operators planning or executing adversary simulation engagements against mature defenders.
Comprehensive lateral movement tradecraft for authorized red team engagements covering credential-based movement (pass-the-hash, pass-the-ticket, overpass-the-hash), NTLM relay attacks (ntlmrelayx with PetitPotam, DFSCoerce, PrinterBug coercion), remote execution protocols (WMI, WinRM, DCOM, PsExec and alternatives), RDP session hijacking, and network pivoting through tunneling tools (chisel, ligolo-ng, SSH tunnels, SOCKS proxies). Provides operator-ready command sequences for mimikatz, crackmapexec/netexec, impacket suite, and evil-winrm with emphasis on OPSEC considerations, SMB signing bypass, and detection evasion. Maps to MITRE ATT&CK T1021 (Remote Services), T1550 (Use Alternate Authentication Material), and sub-techniques. Includes defender-perspective detection guidance for blue team awareness and a rapid engagement cheatsheet for common lateral movement scenarios encountered during internal penetration tests and assumed-breach exercises.
Phishing campaign execution methodology for authorized red team engagements. Covers end-to-end campaign lifecycle: infrastructure provisioning (GoPhish, SMTP relay configuration, domain acquisition and aging, SPF/DKIM/DMARC alignment), payload delivery vectors (Office macro weaponization, HTA droppers, ISO/IMG container abuse, LNK shortcut hijacking, OneNote embedded payloads, HTML smuggling), email authentication bypass techniques (SPF softfail exploitation, DKIM replay attacks, display name spoofing, homoglyph and cousin domain registration), credential harvesting with MFA bypass (EvilGinx2 transparent proxy, Modlishka session relay, pixel-perfect HTML cloning), spear phishing pretext development informed by OSINT, email security gateway evasion, QR code phishing (quishing), and callback phishing for initial access. Integrates with GoPhish for campaign management, EvilGinx2 for adversary-in-the-middle credential interception, King Phisher for template design, and the Social Engineering Toolkit for payload generation. Maps to MITRE ATT&CK T1566 (Phishing) and sub-techniques T1566.001 (Spearphishing Attachment), T1566.002 (Spearphishing Link), T1566.003 (Spearphishing via Service). Assumes you have written authorization and a defined scope before any campaign execution.
Video strategy, scripting, and optimization for YouTube, TikTok, Reels. Use when the user requests 'video strategy', 'YouTube', 'video script', 'Reels', 'video SEO', 'thumbnail', 'video ad', 'explainer video', 'short-form video', or 'video content calendar'.
Complete toolkit for PDF manipulation: text and table extraction, creating new PDFs, merging or splitting documents, and handling forms. Use when Claude needs to fill out a PDF form or process, generate, or analyze PDF documents programmatically at scale
PHPUnit testing best practices and conventions guide. Use when writing, reviewing, or refactoring PHPUnit tests to ensure consistent, maintainable, and effective test suites. Triggers on tasks involving test creation, test refactoring, test configuration, code coverage, data providers, mocking, or PHPUnit XML configuration.
Orchestrate the full Roblox game development lifecycle. Use as the umbrella entrypoint for broad requests like create a Roblox game, build a new game, continue development, what should we do next, take this from idea to launch, make this game profitable, plan the roadmap, manage production, or maintain a live Roblox game. Routes to the right Roblox specialist skills by GDLC phase.
Design, inspect, and refactor Roblox Luau game architecture. Use for ModuleScript/service boundaries, client-server flow, RemoteEvents/RemoteFunctions, ReplicatedStorage organization, reusable systems, script placement, and safe Studio MCP code-edit workflows.
Expert-level Rust development with ownership, lifetimes, async, error handling, and production-grade patterns. Use when the user mentions systems programming, memory safety, concurrency, or Cargo, or when the task involves Ownership and Borrowing, Type System, Async Programming, or Web Development.
Apply and diagnose Rust style, rustfmt, Clippy, compiler diagnostics, Edition migrations, lint policy, idiomatic control flow, error handling, allocation behavior, production Rust conventions, and the Rust API Guidelines ↔ Clippy lint mapping. Use when users ask to format or lint Rust, fix warning or error codes, migrate editions, review unwrap or clone usage, improve idioms, map a C-* API guideline rule to the enforcing Clippy lint, or establish CI quality gates. Route API shape decisions (naming conventions, type/trait design, module layout, full C-* guideline review) to the rust-api-design skill.
(macOS only) Use this skill for ALL Apple Calendar and Reminders management tasks using osascript and reminders-cli, following GTD methodology. Invoke immediately when the user wants to: schedule a meeting or event, add a todo or reminder, check today's agenda, plan the week ahead, do a weekly review, or capture any task they don't want to forget. Also use when the user says 安排会议, 创建提醒, 查看日程, 规划下周, 添加待办, 记一下, 别忘了, 今天要做什么, 周回顾. This is the right skill any time the user wants to interact with macOS Calendar or Reminders apps.
Deep learning for single-cell analysis using scvi-tools. This skill should be used when users need (1) data integration and batch correction with scVI/scANVI, (2) ATAC-seq analysis with PeakVI, (3) CITE-seq multi-modal analysis with totalVI, (4) multiome RNA+ATAC analysis with MultiVI, (5) spatial transcriptomics deconvolution with DestVI, (6) label transfer and reference mapping with scANVI/scArches, (7) RNA velocity with veloVI, or (8) any deep learning-based single-cell method. Triggers include mentions of scVI, scANVI, totalVI, PeakVI, MultiVI, DestVI, veloVI, sysVI, scArches, variational autoencoder, VAE, batch correction, data integration, multi-modal, CITE-seq, multiome, reference mapping, latent space.
Opens an isolated work front in this repository: creates a git worktree outside the project folder, with a new branch based on the remote's UPDATED production branch, prepares the folder (dependencies, local files), and opens it in the editor. On first run, investigates the repository and writes its findings to sdd/ambiente.md. Use when the user says 'open a work front', 'new session', 'new isolated task', 'create a worktree for X', or wants to start a feature/fix without mixing it with another in progress.
Generate, revise, translate, and adapt song lyrics plus compact style prompts for Suno or similar AI music generators. Use when the user wants ready-to-paste AI music generation lyrics, lyric tags, Suno/Udio-style prompts, vocal direction, instrumentation, genre/style wording, song structure, multilingual singable lyrics, or a clarifying interview for an AI-generated song. Do not route general music theory, ordinary artist research, or non-generation songwriting discussion here unless the requested deliverable is lyrics or a generation-ready style prompt.
Breaks requirements down into a structured task list and generates a task management system for long-running Agents (based on Anthropic Effective harnesses methodology). Automatically triggers when users need to manage multi-session development tasks, track feature completion, or request “break down tasks,” “task management,” or “project planning.”
Use when enabling shadows in a Three.js scene or debugging shadow artifacts like acne, peter panning, or low resolution. Prevents the common mistake of wrong shadow camera frustum, too many PointLight shadows, or missing castShadow/receiveShadow flags. Covers shadow map types, per-light config, bias tuning, CameraHelper debugging. Keywords: shadows, shadowMap, shadow acne, peter panning, bias, normalBias, castShadow, receiveShadow, PCFSoftShadowMap, shadow camera, shadows not showing, shadow artifacts, enable shadows.
Use when choosing or configuring materials, loading textures, or setting up PBR workflows in Three.js. Prevents the common mistake of wrong color space on textures, forgetting material.dispose(), or not calling material.needsUpdate after changes. Covers all 15+ material types, PBR metalness/roughness, MeshPhysicalMaterial, texture maps, color space. Keywords: MeshStandardMaterial, MeshPhysicalMaterial, material, texture, PBR, metalness, roughness, normalMap, color space, SRGBColorSpace, blurry texture, pixelated, texture quality, filtering.
Audits a Claude Code setup for token waste and reports only, changing no file or setting. Measures each cost driver with shell and file tools instead of estimating, covering CLAUDE.md sizes and @imports, MCP server and tool counts, whether tool deferral is active, a proxy that silently disables it, model and effort and mid-session switches, output-reducing hooks, per-agent model inheritance, cron and loop intervals versus the measured cache TTL, and the newest session log's token split. Emits one severity-ranked table plus the single highest-leverage fix. Use when the user asks to audit token or context waste, cut token cost, explain a low cache-hit rate, or diagnose context bloat. Triggers include "token waste", "token audit", "context bloat", "cache hit rate", "why is my context so big", "cut token cost".
Polishes speech transcripts (interviews, speeches, podcasts, meetings) into more readable article paragraphs. Use when the user mentions “字幕精修”, “transcript polish”, “润色字幕”, “把视频字幕整理成文章”, “访谈文字整理”, processing interview notes, optimizing transcripts, organizing speech-to-text output, or turning long dialogue or speech transcripts into readable articles. Suitable for transcripts of solo speeches or multi-person conversations; preserves the original sentences and wording and avoids excessive summarization. Use this skill even if the user simply says “帮我整理一下这段文字” and provides clearly colloquial text.
Investigate and retire unused product, code, operational, or infrastructure surfaces using real usage evidence. Use when a repository needs active pruning, when a specific capability may no longer be necessary, or when a scheduled review should preserve retirement candidates without executing them. Do not use for routine lint cleanup, broad refactoring, or dependency updates.
Create a UAT test case suite/checklist for non-IT business users (C&B, accounting, operations)—one offline-proof HTML file with a sequential scenario table searchable by code with Ctrl+F, check/not-check scope by data source, step-by-step instructions with real screenshots blurred for PII, Pass/Fail status, and Excel/CSV/PDF export. Triggers: “create test cases for non-IT,” “UAT checklist for business users,” “test scenarios for C&B,” “/uat-nonit-testcase.”
Use when the user asks to "write a QUnit test", "fix a failing QUnit test", "add a QUnit module", "modernize QUnit tests", "migrate from QUnit 1", or mentions QUnit-specific constructs such as assert.async, nextUIUpdate, Core.applyChanges, sinon sandbox, asyncTest, or QUnit.module. Covers coding standards for OpenUI5/SAPUI5 unit test files: const/let over var, arrow functions over .bind(this), async/await over assert.async(), assert.expect() in every async test, sinon.createSandbox(), descriptive test names, beforeEach/afterEach module isolation, nextUIUpdate vs Core.applyChanges rules, try/finally teardown in helpers, QUnit 1 to QUnit 2 globals migration, and non-ASCII character avoidance.
Writes and reviews native-quality Vietnamese (vi-VN) finance content — hóa đơn điện tử and e-invoice fields, báo cáo tài chính and management reports, investor updates and board decks, pricing and payment-terms pages, fintech, banking and insurance product copy, loan and credit disclosures, and thuế GTGT, TNDN and TNCN tax correspondence. Use when drafting Vietnamese financial statements or financial promotion, applying Thông tư 99/2025 statement terminology, checking guaranteed-return, lãi suất, insurance or digital-asset advertising limits, or formatting statement tables, negatives and đơn vị tính. Flags regulated phrasing and routes it to legal review rather than editing around it.
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.