Skills
101,174 packaged capabilities and playbooks for agents. Install any of them with one command.
All
The lyt series skill for product selection from scratch and building a candidate pool. Covers only cross-border e-commerce, TikTok Shop, e-commerce category selection, product selection boundaries, user capability triage, standardized/non-standard products, supply chain/inventory, product sourcing paths, and candidate product pool building; suited to users who do not yet know what to sell, only know a broad category, want to judge whether a category/country/traffic channel suits them, or want to build a set of candidate products. When the user already has a specific product, candidate product list, product screenshots, 1688/competitor links, supply chain materials, or a candidate product spreadsheet and wants to judge whether it can be done, switch to lyt-product-validation. Triggers: /lyt-product-selection, /选品, /从0选品, /选类目, /建候选品池
Guidance for governing and managing Microsoft 365 collaboration sprawl — Teams/group lifecycle, sharing and guest access governance, and SharePoint Advanced Management — to keep the data estate secure and Copilot-ready. WHEN: Microsoft 365 governance, Teams sprawl, group lifecycle, guest access governance, external sharing controls, SharePoint Advanced Management, site lifecycle, manage collaboration sprawl, container governance.
Convert various file formats (PDF, Office documents, images, audio, web content, structured data) to Markdown optimized for LLM processing. Use when converting documents to markdown, extracting text from PDFs/Office files, transcribing audio, performing OCR on images, extracting YouTube transcripts, or processing batches of files. Supports 20+ formats including DOCX, XLSX, PPTX, PDF, HTML, EPUB, CSV, JSON, images with OCR, and audio with transcription.
Video/audio/image processing with FFmpeg and ImageMagick. Tools: FFmpeg (video/audio), ImageMagick (images). Capabilities: format conversion, encoding (H.264/H.265/VP9/AV1), streaming (HLS/DASH), filters, effects, thumbnails, watermarks, batch processing, hardware acceleration (NVENC/QSV). Actions: convert, encode, resize, crop, compress, extract, merge, stream, transcode media. Keywords: FFmpeg, ImageMagick, video encoding, audio extraction, image resize, thumbnail, watermark, HLS, DASH, H.264, H.265, VP9, AV1, codec, bitrate, framerate, resolution, aspect ratio, filter, overlay, concat, trim, fade, batch processing. Use when: converting video/audio formats, encoding with specific codecs, generating thumbnails, creating streaming manifests, extracting audio from video, batch processing images, adding watermarks, optimizing file sizes.
Custom middleware patterns for ASP.NET Core applications. Covers request/response pipeline, middleware ordering, conditional middleware, IMiddleware factory pattern, IExceptionHandler (.NET 8+), and reusable middleware components. Use when creating custom middleware in ASP.NET Core applications, understanding middleware pipeline ordering, or implementing cross-cutting concerns like logging, authentication, and caching.
MLA (Multi-Latent Attention) cost models, regime analysis, and kernel selection guide. Use when: (1) reasoning about which kernel approach to use for a given regime, (2) understanding cost model tradeoffs between FlashMLA, FlashAttention, and MLAvar6+, (3) analyzing roofline behavior across decode/speculative/prefill regimes, (4) setting optimization targets, (5) understanding MLA math and absorption trick.
moomoo CLI market data skill for stock, option, futures, quote, K-line, snapshot, ticker, order book, ranking, capital flow, company, financials, shareholders, valuation, plate, IPO, calendar, watchlist, and screening quote commands; not for trading orders
Split a large task across parallel sub-agents with clean role boundaries and a merge step. Use when the user asks to "parallelize", "use multiple agents", "fan out", or when a task has independent workstreams that bottleneck on one agent.
Design accessible video, audio, and multimedia content with captions, transcripts, and audio descriptions. Use when creating or reviewing video, audio, podcasts, webinars, animations, or any time-based media. Triggers on: captions, subtitles, transcript, audio description, video accessibility, podcast accessibility, multimedia, closed captions, sign language, media player, autoplay.
This skill should be used when crafting prompts for Nano Banana Pro (Gemini image generation). Use when users want help writing image generation prompts, need guidance on prompt structure, or want to optimize their prompts for better results.
Queries the NCATS Translator ARAX production API for bounded, typed, provenance-rich one-hop and endpoint-pinned two-hop biomedical knowledge-graph relationships. Use for Biolink-constrained RTX-KG2 lookup, explicit selected-provider ARAX federation, separate entity normalization, qualifier-aware graph traversal, and inspection of TRAPI edge bindings, publications, and knowledge-source provenance. Do not use for inference, ranking, open-ended pathfinding, clinical guidance, or sensitive queries
Evaluates LLMs across 100+ benchmarks from 18+ harnesses (MMLU, HumanEval, GSM8K, safety, VLM) with multi-backend execution. Use when needing scalable evaluation on local Docker, Slurm HPC, or cloud platforms. NVIDIA's enterprise-grade platform with container-first architecture for reproducible benchmarking.
Complete Next.js deployment system (Next.js 15.5/16). PROACTIVELY activate for: (1) Vercel deployment and configuration, (2) Self-hosted Node.js deployment, (3) Docker containerization, (4) Static export with output:'export', (5) Edge Runtime configuration, (6) Environment variables, (7) CI/CD with GitHub Actions, (8) Health checks and monitoring, (9) Production optimization, (10) Turbopack production builds (Next.js 16). Provides: Dockerfile, docker-compose, PM2 config, vercel.json, GitHub Actions workflows, Turbopack config. Ensures production-ready deployment with proper configuration.
Complete Next.js Middleware system (Next.js 15.5/16). PROACTIVELY activate for: (1) Creating middleware.ts (Edge) or proxy.ts (Node.js), (2) Matcher configuration, (3) Authentication protection, (4) Role-based access control, (5) Redirects and rewrites, (6) Internationalization (i18n), (7) Security headers (CSP, CORS), (8) Geolocation routing, (9) Bot detection, (10) Rate limiting patterns, (11) Node.js Middleware with proxy.ts (Next.js 16). Provides: Edge middleware patterns, Node.js proxy patterns, matcher syntax, auth guards, header manipulation, cookie handling, database access in proxy.ts. Ensures optimized request handling with proper security.
Complete Next.js Server Actions system (Next.js 15.5/16). PROACTIVELY activate for: (1) Defining Server Actions with 'use server', (2) Form handling with actions, (3) useActionState for form state, (4) useFormStatus for pending states, (5) Optimistic updates with useOptimistic, (6) Validation with Zod and next-safe-action, (7) Authorization in actions, (8) File uploads, (9) Revalidation after mutations, (10) Type-safe actions with next-safe-action library. Provides: Action patterns, form integration, validation, optimistic UI, error handling, next-safe-action integration. Ensures secure server mutations with proper validation and UX.
Provides guidance for interpreting and manipulating neural network internals using nnsight with optional NDIF remote execution. Use when needing to run interpretability experiments on massive models (70B+) without local GPU resources, or when working with any PyTorch architecture.
Medical image segmentation with nnU-Net's self-configuring framework — auto-selects architecture, preprocessing, training for any modality. CT, MRI, microscopy, ultrasound in 2D, 3D full-res, 3D low-res, cascade. Pipeline: convert → plan/preprocess → train (5-fold CV) → best config → predict → ensemble. Use when classical segmentation fails and annotated data exists.
Analyze Northbound capital (Shanghai Connect/Shenzhen Connect) flows, sector preferences, divergence signals, and risk alerts. Use when the user asks about Northbound capital, foreign capital inflows and outflows, Northbound preferences, or needs Northbound capital monitoring.
Operate Notion through the notion-axi CLI — search; read, create, update, archive, and move pages; create and edit databases and query/set typed row properties; manage blocks, comments, and file uploads; or call any REST endpoint directly. Use whenever a task touches Notion: finding pages or databases, reading or editing content, working with database rows, comments, files, or anything else in a Notion workspace.
Mandatory pre-code gate and definition-of-done for ANY Odoo change (add field, override method, inherit view/xpath, OWL/JS patch, wizard, cron, controller, report, security, migration, bug fix, refactor, "optimize", "clean code", "stick to base"), Odoo 16-19. Also when the user pastes an Odoo traceback or install/upgrade error (ParseError "cannot be located in parent view", "Field ... does not exist", missing xmlid) or asks where an Odoo model, field, method, view or xmlid is defined. Run BEFORE brainstorming or planning: Odoo design questions are answered by tracing the real source into a Context Brief (file:line), not by open-ended Q&A or memory. Resolves the Odoo runtime and version, loads the matching odoo-<major> pack, and closes with install, test, upgrade, i18n, security and hygiene checks before commit.
Automate Gitcode Pull Request creation workflow for OpenHarmony graphic subsystem including code commit, issue creation, and PR submission with proper template formatting. Use when user needs to create a PR for the Gitcode platform, especially for OpenHarmony graphic projects that require specific PR templates with CodeCheck tables and Signed-off-by tags.
Use when a project has NO okteto.yaml/okteto.yml and the user wants it on Okteto — e.g. "get this repo onto Okteto", "set up a dev environment for this", "create an Okteto manifest", "onboard this service". Also use when okteto deploy fails because no manifest exists. Do NOT use if okteto.yaml or okteto.yml already exists — that is the okteto skill's domain.
Computer vision for bio-image preprocessing, feature detection, real-time microscopy. Color conversion, morphology, contour/blob detection, template matching, optical flow on fluorescence/brightfield. 10-100× faster than pure Python via C++. Use scikit-image for scientific morphometry/regionprops; OpenCV for real-time, video, classical feature extraction.
Use when communicating, deciding, or reporting as a senior operator — 10 behavioral modules from Anthropic's Fable 5 and Fable 5.1 prompting guides. Not capability — pure behavior. Load for user-facing text quality, narrating long work, and keeping edits in scope.
YOLO mode. Spawns 4 parallel C-suite agents (CEO, CTO, CFO, COO). Each analyzes the business from their perspective using ALL available data. Produces unfiltered Hard Truths report. After user types YOLO, autonomously runs the business for a day using /loop.
When the user wants to build optimization models, solve mathematical programming problems, or find optimal solutions. Also use when the user mentions "linear programming," "integer programming," "mixed-integer programming," "constraint optimization," "mathematical optimization," "operations research," "optimal solution," or "decision optimization." For network design specifically, see network-design. For routing, see route-optimization.
Use the public `orca` CLI to operate Orca-managed worktrees, folder contexts, terminals, repos, automations, artifacts, worktree comments, and the browser embedded inside the Orca app. Use when the user says "$orca-cli", "use orca cli", "Orca worktree", "child worktree", "cardStatus", "spawn codex/claude in a worktree", "read/wait/send Orca terminal", "terminal send", "full handoff", "handover", "give this to another agent", "another worktree", "Orca browser", "orca artifacts", "share HTML/Markdown", "public artifact link", or "control the browser inside Orca". Prefer this over raw `git worktree`, ad hoc PTYs, Playwright, or Computer Use when the task touches Orca-managed state. Use Computer Use for browser windows, webviews, or desktop UI outside Orca's embedded browser.
Pipeline skill for automating prior authorization decision workflows. Use when the user asks to parse PA request data (X12 278 or FHIR PAS bundles), extract clinical features for adjudication, build rules-based PA decision engines, train ML classifiers on historical PA decisions, analyze denial patterns, or generate SHAP explanations for PA outcomes. Triggers include "parse 278", "FHIR PAS bundle", "PA automation", "adjudication logic", "PA classifier", "denial analysis", "prior auth ML", "SHAP explainability", "PA feature extraction", "rules engine PA", "PA decision pipeline", "authorization workflow", "clinical criteria extraction", "denial pattern mining", "PA turnaround time".
Use when evaluating an Armory package, auditing agent or hook quality, checking package completion contracts, or reviewing the catalog across package types. Covers single-package and repository-wide audits. Not for LLM prompt optimization or ordinary PR code review.
Paper interpretation assistant. Use when the user sends an arXiv paper link and wants to download the paper, interpret it, generate reading notes, break down the paper, or output a detailed report. It creates a paper folder in the working directory, downloads the PDF and TeX Source (if available), and generates a Chinese Markdown report. By default it delivers a first draft and does not review automatically; follow-up review is arranged only if the user explicitly agrees. Not for cases where only a short recommendation is needed.
Fine-tunes LLMs with Hugging Face PEFT, using LoRA, QLoRA, IA3, AdaLoRA, prefix tuning, and prompt tuning so that under 1% of parameters are trained. Covers rank, alpha, and target module selection, loading and merging adapters, multi-adapter serving, and integration with TRL SFTTrainer, Axolotl, and vLLM. Use when fine-tuning 7B-70B models on limited GPU memory, when running QLoRA on a single 24GB GPU, when choosing LoRA rank and alpha, when merging or swapping adapters on one base model, or when debugging CUDA OOM or an adapter that does not apply. Not for full fine-tuning of small models under 1B parameters or cases needing all weights updated.
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.