Skills
97,501 packaged capabilities and playbooks for agents. Install any of them with one command.
All
Manage Vikunja tasks via VJA CLI, supporting CRUD operations, assignment, priority, and labeling. Use for task-related requests like creating, listing, updating, or deleting tasks. Always append --json to commands for structured results, and pipe to jq for compact and filtered output where possible.
Use when surveying the DOCUMENTED visual and interaction conventions of a product's domain BEFORE any wireframe, design system or hi-fi screen is produced — deriving a UI-pattern vocabulary map, executing ONE search angle across design-system documentation, the ARIA Authoring Practices Guide, WCAG success criteria, platform human-interface guidelines and the deceptive-pattern corpus, deep-reading ONE convention source into a record, or synthesising the convention register and report the downstream design skill consumes. Mines documentation; live-site captures are observed, never prescribed. Produces schema-validated artifacts whose coverage grid records every query as run, so a domain with no documented convention is distinguishable from a search that never ran. Carries a design system's DTCG tokens verbatim, never blended. Keywords: visual prior art, design system research, UI patterns, interaction conventions, accessibility criteria, design tokens, dark patterns.
Structured visual QA verdict — compare generated UI screenshots against one or more reference images and return a strict JSON verdict that can drive the next round of changes. Use for screenshot fidelity, layout spacing, font, and color consistency acceptance (trigger words: 对比截图, 视觉验收, 还原度, UI 截图比对; visualize / screenshot / visual verdict).
Render inline interactive visualizations directly in the conversation — SVG diagrams, HTML components, charts, explainer graphics. Trigger when the user asks to draw, make a diagram, draw a flowchart, architecture diagram, data visualization, UI mockup, or says "draw a picture", "show me", "draw", "map out", "visualize", "diagram"; even if not stated explicitly, use it proactively whenever the topic involves spatial, sequential, or systemic relationships that a diagram makes clearer than text.
Expert voxel artist with deep knowledge of MagicaVoxel workflows, palette design, animation techniques, mesh conversion, and game engine optimization. This skill represents insights from professional voxel artists, Teardown developers, and the MagicaVoxel community - knowledge Claude wouldn't normally have about this specialized art form. Use when "voxel, voxel art, magicavoxel, qubicle, voxel model, voxel animation, voxel game, blocky art, minecraft style, teardown style, voxel character, voxel environment, voxel optimization, voxel export, voxel to mesh, marching cubes, greedy meshing, voxel palette, destructible voxels, voxel, 3d, pixel-art, game-dev, magicavoxel, qubicle, animation, optimization, indie-games, retro, blocky" mentioned.
Manages the walletconnect CLI for wallet connection, session management, message signing, transaction sending, and cross-chain bridging (swidge). Use when the user wants to connect a wallet, check session status, sign messages, send transactions, bridge tokens across chains, or disconnect.
Use when working with Outpost Bio's open microbiome foundation models - the Waypoint checkpoints (Waypoint-6m, Waypoint-45m, Waypoint-170m), the Atlas pretraining corpus, the Compass eight-task benchmark, or the `waypoint` CLI from the `waypoint-bio` package. Covers embedding microbiome samples, fine-tuning on taxonomic abundance data, benchmarking a checkpoint on Compass, pretraining a GPT-2 model on taxonomic abundance profiles, and converting MetaPhlAn, Kraken2, QIIME 2, or MGnify abundance tables into waypoint format.
Smart contract security audit — 10 DeFi bug classes (accounting desync, access control, incomplete path, off-by-one, oracle, ERC4626, reentrancy, flash loan, signature replay, proxy), pre-dive kill signals (TVL < $500K etc), Foundry PoC template, grep patterns for each class, and real Immunefi paid examples. Use for any Solidity/Rust contract audit or when deciding whether a DeFi target is worth hunting.
This skill helps users extract full article contents from WeChat using the BrowserAct API. The Agent should proactively apply this skill when users express needs like finding full WeChat articles for specific keywords, tracking WeChat public accounts for industry trends, extracting WeChat article contents for media research, monitoring public relations on WeChat platforms, collecting competitor updates from WeChat, getting full article body from WeChat links, monitoring brand exposure on WeChat articles, retrieving structured WeChat data for sentiment analysis, summarizing daily news from WeChat, getting author and publication date for WeChat articles, or automating WeChat content extraction without scraping.
Automatically search, fetch and rewrite technical content to produce Chinese popular-science articles suitable for WeChat official accounts. Covers AI LLMs, GitHub open source tools and technical topics. Use when the user says "write an article about XXX", "help me research XXX" or "generate a WeChat official account article".
Builds a product wiki from the application source code or updates it after large code changes (release, refactor, many merged PRs). Records mismatches between code and claims about it (comments, docstrings, developer docs, API specs, configs). Use when the user asks to build or update the wiki/requirements/specification/docs from code.
AI/ML integration reference for Windows app development (Windows AI Foundry). Covers Windows AI APIs, LanguageModel, TextRecognizer, ImageScaler, ImageObjectExtractor, ImageObjectRemover, ImageDescriptionGenerator, SemanticSearch, ContentModeration, Phi Silica / Aion Instruct, LoRA, Windows ML, Windows ML Runtime API, ONNX Runtime, execution providers, WebGPU EP, Model Catalog, LearningModel, Foundry Local (CLI / REST API / SDK), MCP servers, App Actions, IActionProvider, Agent Launchers, DirectML, IDMLDevice, operators, tensors, and NPU.
WOLVERINE v6.0.0 — HYPE Patient-Conviction. Six-gate HYPE single-asset thesis preserved, behavior shifts to fewer-stronger entries with multi-day DSL holds (Bison v3.0.1 pattern port). MIN_SCORE 10, max 2 entries/day, 4h same-asset cooldown, Phase 2 ladder widened to T0 lock 0 so a +10% mover can retrace fully without locking — the structural change that lets winners run into BIG trends instead of getting clipped at the first ratchet. Producer adds recent-signals race-window dedup that eliminated ENGINE_FAILURE retry noise on Bison's decision log. Conviction-tiered leverage (3x standard / 5x apex) preserved — HYPE volatility doesn't yet warrant Bison's 7-10x majors-tier sizing.
Use when the user says 'plan a campaign', 'promote this product', 'write the copy for our sale', 'what should we push this month', or 'draft the launch for this collection'. Reads what the store actually sells, drafts the campaign around it, and stages every page and product change for approval.
Produces step-by-step computational walkthroughs of vector and matrix operations as a sequence of numbered "frames", showing the explicit state at each step. The text-equivalent of a 3Blue1Brown animation — each frame shows what changed and why, so the learner can re-trace the operation by hand. Use when the learner needs to *see* a computation unfold (eigenvalue computation, attention with 3 tokens, gradient descent step, SVD on a 2×2, layer norm on a 3-vector, softmax of a small input), when an explanation has been given but the learner needs to ground it in a worked example, or when introducing an operation that's intimidating in symbol form but trivial in pencil-and-paper form.
Use PostHog Autoresearch, agent-driven predictive models: ask "who will do X in the next N days?", let an agent train a model, then score people daily into a person property and `autoresearch_prediction` events. Use when a user wants to predict churn, activity, return, feature adoption, conversion or any event or action per person; create or validate a model from a template; start training or score now; read model quality (holdout vs realized AUC, lift, calibration, likelihood segments, coverage); steer training with suggestions; act on predictions with cohorts, feature flags, workflows or SQL; or diagnose a model that fails to train, scores everyone the same, or stops scoring. Trigger on "autoresearch", "predictive model", "propensity score", "churn prediction", "likely to convert", "who will do X", "predicted_p_", "autoresearch_prediction", "realized AUC", "champion model".
Complete reference for WoW Retail Currency, Auction House, Token, Store, Trading Post, Black Market, Mail, and Player Trading APIs. Covers C_CurrencyInfo (currency data, tracking, backpack), Auction House (full listing/bidding/buying/selling/search/commodity), C_WowTokenPublic (WoW Token market), AccountStore (in-game shop), CatalogShop, C_PerksProgram (Trading Post — Trader's Tender), BlackMarket (BMAH), TradeInfo (player trading), MailInfo (mailbox), and related events. Use when working with currencies, auction house UI, WoW Tokens, the in-game store, Trading Post, black market, mail, or player-to-player trading.
Official document writing assistant for China's government system (government style). Covers statutory official documents (notice/request for instructions/report/reply/letter/opinion/bulletin/briefing/decision/order/announcement/public notice/communiqué/motion/resolution) and transactional materials (leadership speech drafts/exchange remarks/work summaries/key work points/work plans/reporting materials/research reports/government information/statement of circumstances/self-assessment reports). Trigger scenarios: "help me write a ×× (any of the above document types)", "turn these key points/meeting notes into an official document", "how do I handle official document formatting/rules", "revise/polish/review an official document", "export an official document to Word/docx".
Turn agreed product, domain, scope, and architecture decisions into one traceable, implementation-ready feature specification with observable acceptance criteria and verification points. Use when a requirement is sufficiently decided to specify behavior but should not yet be decomposed into implementation tasks.
Create engaging Substack Notes that build community and grow subscribers. Use this skill when asked to write Substack Notes, short-form content for Substack, or help with Substack growth strategy. Triggers include requests for Substack Notes, Note ideas, or Substack engagement content.
X/Twitter public sentiment research. Searches X for real-time perspectives, market sentiment, expert opinions, breaking news, and community discourse. Use when: user asks "what are people saying about", "X/Twitter sentiment", "check X for", "search twitter for", "what's CT saying about", or wants public opinion on a stock, sector, company, or market event.
An AI cover design tool built for Xiaohongshu content creation. It draws on 2,000+ viral post data points collected daily from across the web, analyzes visual elements of top-performing covers in the same niche, and uses AI to identify visual patterns that drive conversions. It generates actionable cover design plans tailored to post content and platform aesthetics, saving tedious design work and increasing cover click-through rates.
Create Xiaohongshu (RED / 小红书) note covers and first images in 3:4 (1080×1440) for product seeding (种草), reviews, hauls and tutorials: big bold Chinese headline, product hero, visual anchor and stickers, in proven viral layouts. Use for 小红书封面, 种草图, 笔记首图, RED cover, rednote.
Extends Fastly VCL with loops, functions, constants, macros, conditionals, and includes via XVCL — a VCL transpiler that compiles .xvcl files into standard VCL. Use when writing VCL for Fastly, working with .xvcl files, generating repetitive VCL (multiple backends, routing rules, headers) with loops, defining reusable VCL functions with return values, using compile-time constants instead of magic numbers, or writing any Fastly VCL configuration. XVCL syntax is not in training data so this skill is required. Also applies when writing and testing VCL locally (compile with `uvx xvcl`, test with falco), reducing VCL code duplication, splitting large VCL into modular includes, or doing any VCL development task for Fastly — even without explicitly mentioning XVCL.
【XY Operations System】The main entry point to 风物执守 (XY), a business and private-domain operations system (private-domain business at its core, with short video/traffic/AI as built-in capabilities). Provides beginner tutorials, pre-task routing, and post-task navigation. Use when the user does not know which xy Skill to use, describes a business/account/content/team problem, or asks “what should I do next?”｜小爷出品
【AI pipeline】An AI pipeline for content production. First audit who handles each step from idea to publishing and where work gets stuck; then decide which steps should go to AI and which must remain human. Orchestrate idea → topic → transcript → clips → title → publishing and troubleshooting into a pipeline with handoff artifacts. Teach the method of “turning a step into a skill once AI can run it reliably,” and set ethical boundaries for style imitation (copy the structure, not someone else’s writing). Trigger when the user says “I want to build an AI content pipeline,” “which steps can AI do,” “how do I turn this step into a skill,” “how can I make AI write like me,” or “my content production is too slow.”｜小爷出品
【Archive】Save the conclusions, rejected paths, and untested assumptions from this discussion in a local archive for next time, and manage where archives are stored. Use when the user says “keep this conclusion,” “save this,” “note this down,” “save this diagnosis,” “where are the archives,” or “change the archive location.” Trigger: /xy-archive, “save this,” “remember this,” “where are my archives”｜小爷出品
【Atomic ingestion】Extract local content assets into XY atoms for ingestion: specify a directory (manuscripts/transcripts/posts/course materials), extract them to the atomic standard as jsonl in incoming/: audit first, prepare 3 sample pieces for user approval, then process in batch, deduplicate, verify, and finally prompt to ingest with build_atoms. Use when the user says “turn these drafts into atoms,” “ingest my transcripts,” or “extract insights in bulk.”｜小爷出品
【Business diagnosis】Business model diagnosis. Two modes: consultation (first break down the problem you bring in, assess whether it holds up, then decide whether to answer) and checkup (review your business across XY’s seven checks, from net profit to growth stage, and produce a report). Trigger when the user says “take a look at my business,” “diagnose my model,” “I have a business question,” “what’s wrong with my business,” “should I keep going,” or “help me figure out whether I’m making money.”｜小爷出品
【Consolidated report】Combine multiple xy-archive records collected under a project into a chronological markdown diagnostic report that can be shared with partners, archived, and revisited three months later. Use when the user says “make a report,” “package these up,” “organize this into something I can show my partners,” or “combine the conclusions from these sessions.” Trigger: /xy-brief, “make me a report,” “package these up”｜小爷出品
【Decision review case file】Turn a domain that needs long-term tracking—a private-domain business, an account matrix, a product or business-model change—into a local, four-layer knowledge project: facts, patterns, snapshots, and open questions. Include source labels and immutable written snapshots; allow outcomes to be backfilled and your own patterns to be distilled. Use when the user says “track this decision,” “open a case,” “the results are in; backfill them,” “make me a status profile,” or “here are the recent changes…”. Trigger: /xy-casefile, /决策立案, /结果回填, /状态画像｜小爷出品
【Long-form content slicing】Turn transcripts of courses, livestream reviews, interviews, and private-domain training into short-video plans that can each be published on their own. The opening, development, and ending may come from three distant points in the transcript, jump-cut together into a complete narrative. Output plans only; do not add a single word. Use when the user says “slice this for me,” “how many clips can we cut from this livestream,” “here’s the transcript; slice it,” “course clips,” or “find a few sections I can post.”｜小爷出品
【Sales conversion diagnosis】Diagnose private-domain sales conversion, from a customer adding you on WeChat and showing interest through quoting, closing, repeat purchases, upselling, and complaints. Check each stage to find why conversion is stuck: whether the pacing is right, whether there is an anchor, how to structure the first, second, and third conversions, whether follow-up is segmented, and whether you are falling into traps such as low-to-high conversion, 9.9 lead-generation courses, or constant promotions. Trigger when the user says “the customer isn’t replying,” “we talked but they didn’t order,” “how should I quote,” “how do I close,” “repeat purchases aren’t increasing,” “the customer thinks it’s too expensive,” “the customer deleted me,” or “review my sales script.”｜小爷出品
【Content diagnosis】Diagnose content after the topic has been chosen: determine its format and platform, then assess it across five dimensions (writing and authenticity / cover and title / communication efficiency / cognitive gap / fit with content purpose). Produce a report and the next first step. Diagnose only; do not write on the user’s behalf. Trigger when the user says “how should I make this content,” “review this copy,” “what format should I use for this topic,” “review my spoken script,” “why is nobody watching this video,” or “should I post this on Douyin or Xiaohongshu?”｜小爷出品
【Interactive course】Turn a topic into a connected series of learning articles, with a checkpoint at the end of each to collect answers and user feedback; learners cannot proceed until they understand the material. It can also infer what to learn from a piece of good content provided by the user, generate a staged outline, and build a course. For topics in private-domain business, short video, or AI business, use the atomic library directly as course material and include IDs. Trigger when the user says “teach me private-domain sales systematically,” “help me understand direct sales vs. agency,” “continue with the next article,” “write the next one based on my feedback,” “make me a course,” or “what should I learn behind this content?” Trigger words: /xy-course, /xy-learn, /交互式学习.｜小爷出品
【Resonance diagnosis】Diagnose a draft’s resonance and decode its spread in two modes: ① For an unpublished draft, identify the core mechanism using five communication theories, determine which dimension is missing, and suggest cuts or edits if the user fears it will not resonate or get shared. ② For content that has already taken off (the user’s own or a reference), analyze whose thoughts it voiced and why people shared it. Trigger when the user says “does this draft resonate,” “does it hit home,” “why did this take off,” “what do viewers really want to hear,” “what emotion did this hit,” or “is it ready to post?”｜小爷出品
【Goal audit card】Audit goals. Turn aspirations such as “I want to build a private-domain business,” “I want to build an IP,” or “I want to make my team stronger” into actionable, verifiable goal cards (deliverable / for whom / scope / completion criteria). First extract what is already known from the conversation and fill only one gap that would change the approach or acceptance criteria; if things are clear, proceed directly. Use when the user says “I want to do … but don’t know where to start,” “my goal is to become …,” “I want to become more …,” “is this task clear enough?” or “help me set a goal.”｜小爷出品
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.