Skills
100,344 packaged capabilities and playbooks for agents. Install any of them with one command.
All
When the user wants to plan, design, or implement an A/B test or experiment. Also use when the user mentions "A/B test," "split test," "experiment," "test this change," "variant copy," "multivariate test," "hypothesis," "conversion experiment," "statistical significance," or "test this." For tracking implementation, see analytics-tracking.
Draft or revise the abstract and introduction of a research paper. Use whenever the user needs to write or improve these sections — including phrasings like "write my abstract", "improve my intro", "draft the introduction", "my abstract is weak", "how should I open the paper", "the intro is too long", "I need a hook", "write a one-paragraph summary", or "my introduction doesn't sell the contribution". Also use when the user has a paper plan or results but no prose yet for the opening. The abstract and introduction have distinct structures and reviewer conventions that differ from the rest of the paper; this skill handles both as a pair since the intro must match and extend the abstract's promises. If the contribution or paper spine is not yet stable, use paper-argument-planner first.
Plan, generate, inspect, and refine academic figures from repositories, papers, draft notes, paper URLs, PDFs, or reference images. Design evidence-grounded academic figures and construct, diagnose, or revise scientific image prompts, including semantic palettes, reference-led styles, prompt engineering, information-density feedback, and FigureSpec v1 compilation. Supports fast-track draft-to-figure generation, user passthrough mode, and replacing AI-rendered text with editable PowerPoint text boxes.
Load PROACTIVELY when task involves accessibility compliance or inclusive design. Use when user says "check accessibility", "audit for WCAG", "fix screen reader issues", "add keyboard navigation", or "check color contrast". Covers WCAG 2.1 AA compliance across semantic HTML, ARIA patterns, keyboard navigation, screen reader support, color contrast ratios, form accessibility, media alternatives, and focus management. Produces structured audit reports with severity ratings.
Comprehensive writing guide for Angular documentation (adev). Covers Google Technical Writing standards, Angular-specific markdown extensions, code blocks, and components. You MUST use this skill any time you plan to create, edit, or review documentation files in `adev/` or `adev/src/content`.
Use when admin/CMS/back-office screens need to be built or specified after admin-ui-orchestrator defines roles, permissions, workflows, and safety gates. Covers tables, forms, editors, moderation queues, dashboards, CRUD surfaces, and state coverage without production destructive actions.
Intercept the AG2 agent loop with `BaseMiddleware` — wrap full turns (`on_turn`), each LLM call (`on_llm_call`), each tool execution (`on_tool_execution`), or each human-input request (`on_human_input`). Use for retry, logging, history trimming, request mutation, tool auditing, guardrails, or rate limiting. Built-ins: `LoggingMiddleware`, `RetryMiddleware`, `HistoryLimiter`, `TokenLimiter`, `TelemetryMiddleware` (see `ag2-telemetry`). For per-tool hooks see also `ag2-add-custom-tool` tool-middleware section.
Open an AG2 network `discussion` channel — N-party (2+) round-robin where each participant speaks in fixed order, cycling until explicit close or TTL. Use when the user wants a brainstorm with a fixed cast, a panel discussion, or round-robin reviewers. Covers `agent_client.open(type="discussion", target=[...], knobs={"ordering": ORDERING_ROUND_ROBIN})`, the `expected_next_speaker` rotation, the `hc.can_send(...)` probe pattern (default handlers skip LLM calls when it isn't their turn), `DiscussionState`, the `turn_within` expectation defaults (`warn` at 120s / `hide` at 600s), view-window sizing for N participants, and the four close patterns that work with this adapter. Load this after `ag2-network-quickstart`. For conditional handoffs or declarative orchestration, see `ag2-network-workflow` instead.
Use when an agent keeps relearning the same correction, CLAUDE.md has grown unreadable, or you need to trace where a remembered rule came from. Promotes facts through four tiers only on recurrence across sessions and days, with memory_promote.py and a human adopt step.
Agent design advisor and review tool. Based on 12-Factor AgentOps best practices, for: (1) discussing Agent architecture designs; (2) reviewing the design of existing Agents/Skills/workflows, finding issues and giving improvement suggestions. Trigger words: review my agent, help me analyze this skill, agent design, agent optimization, help me review this workflow, what's wrong with this agent, how to design an agent, agent architecture consulting.
Identifies and removes signs of AI-generated text to make writing sound more natural and human. Use when the user asks to humanize text, make writing sound more natural, remove AI patterns, edit for human voice, or convert AI-generated content to natural prose. Also triggers on requests to check for AI writing patterns, review text for robotic language, or improve writing authenticity. Based on Wikipedia's "Signs of AI writing" guidelines.
Distill books, documents, folders, or URLs into compact, practical Agent Skills. Use when source material should become either one reusable skill package or a split set of focused skills, each with a concise SKILL.md plus detailed references and examples.
Algolia data modeling and indexing guidance. Use before or alongside indexing records or building Algolia search UI for net-new search, browse, autocomplete, ecommerce, personalization, Dynamic Re-Ranking, recommendations, or analytics-aware implementations. Also use to audit or review an EXISTING implementation's record shape and data contract. Makes record shape, objectID, display fields, facets, ranking fields, and event attribution explicit decisions. Use for records, variants, SKUs, indices, replicas, searchable and faceting attributes, denormalization, merchandising fields, timestamps, inventory, event attribution, indexing pipelines, partial updates, secured data, multi-language or multi-region strategies, and migrations. Do NOT use for live imports, exports, record mutations, settings changes, or account actions; use algolia-cli or algolia-mcp. Do NOT use for frontend UI implementation; use algolia-instantsearch-ui, algolia-autocomplete, algolia-ui-libraries, or the official instantsearch skill.
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.
Query SAG (Smart Access Gateway / 智能接入网关) configurations and perform status inspections via Alibaba Cloud OpenAPI. Generates inspection report files. Use when the user mentions SAG, smart access gateway, 智能接入网关, or asks to check SAG instance status, query SAG configs, inspect SAG health, troubleshoot SAG connectivity, perform SAG device inspections, 网关巡检, 查询SAG配置, SAG状态巡检, or SAG健康检查.
Use when managing Alibaba Cloud AIContent (AiContent) via OpenAPI/SDK, including the user needs AI content generation or content workflow operations in Alibaba Cloud, including listing assets, creating/updating generation configurations, checking task status, or troubleshooting failed content jobs.
Use when managing and troubleshoot Alibaba Cloud ALB (Application Load Balancer), including the user asks to inspect, create, change, or debug ALB instances, listeners, server groups, rules, certificates, ACLs, security policies, or health checks in Alibaba Cloud.
Use when routing Alibaba Cloud Model Studio requests to the right local skill (Qwen text, coder, deep research, image, video, audio, search and multimodal skills). Use when the user asks for Model Studio without specifying a capability.
Use when managing Alibaba Cloud ApsaraVideo for Media Processing (MPS/MTS) resources and workflows via OpenAPI/SDK, including media ingest and metadata tasks, transcoding/snapshot jobs, pipeline/template/workflow operations, and MPS job troubleshooting.
Use when generating videos with Alibaba Cloud Model Studio PixVerse models (`pixverse/pixverse-v5.6-t2v`, `pixverse/pixverse-v5.6-it2v`, `pixverse/pixverse-v5.6-kf2v`, `pixverse/pixverse-v5.6-r2v`). Use when building non-Wan text-to-video, first-frame image-to-video, keyframe-to-video, or multi-image reference-to-video workflows on Model Studio.
Enforces pre-registration discipline with the Iron Law NO DATA ANALYSIS WITHOUT A PRE-REGISTERED ANALYSIS PLAN FIRST, a spirit-vs-letter line, an Excuse-vs-Reality rationalization table, and a Red-Flags-STOP list (HARKing, optional stopping, post-hoc covariates, outlier-dropping, test-shopping). Runs a PLAN/COLLECT/CONFIRM/EXPLORE workflow that freezes hypotheses, tests, exclusions, and stopping rules before data, then forces unplanned findings to be labeled exploratory (their p-values lose confirmatory status, per COS confirmatory/exploratory model). Orchestrates, not replaces, alterlab-open-science (OSF/AsPredicted registration), alterlab-statistical-analysis (test selection, assumptions), and alterlab-scientific-thinking (bias grading). Use when analyzing data without a frozen plan, switching the primary outcome or adding covariates after seeing results, weighing early stopping, dropping outliers post-hoc, pre-registering a study, or rationalizing deviation. Part of the AlterLab Academic Skills suite.
Enforces results-reporting transparency as a discipline gate built on the Iron Law "NO RESULTS CLAIM WITHOUT REPORTING EVERY ANALYSIS RUN" — a numbered Gate Function (IDENTIFY the claim, LIST every test actually run including the ones that did not "work", CHECK assumptions were reported, CHECK effect size with 95% CI is present, CHECK pre-registration deviations are disclosed, ONLY THEN write the sentence), plus an Excuse-vs-Reality table and Red-Flags-STOP list for selective reporting, cherry-picking, and bare p-values. Use when writing up Results, claiming a finding from a subset of analyses, reporting a p-value without an effect size or confidence interval, dropping outliers post hoc, or omitting analyses that did not pan out. Orchestrates alterlab-statistical-analysis (tests, effect sizes), alterlab-preregistration-discipline (the frozen plan), and alterlab-open-science (TOP, disclosure); it does not run the tests itself. Part of the AlterLab Academic Skills suite.
Quantifies bulk RNA-seq transcript abundance with salmon 2.x (the Rust rewrite; selective alignment or --sketch) and kallisto (v0.52.0, kb-python workflow), builds a decoy-aware gentrome index, runs quant with --gcBias -l A, then imports estimates via tximport/tximeta with a tx2gene map and hands differential expression to alterlab-pydeseq2. Warns that salmon 2.0 cannot read C++/pufferfish indices (rebuild every index), that --validateMappings is now accepted-but-ignored, and that 'salmon alevin' was REMOVED (single-cell now uses piscem + alevin-fry). Use when quantifying RNA-seq transcript abundance, running salmon or kallisto, building a decoy-aware index, or wiring tximport to DESeq2; for differential expression use alterlab-pydeseq2, for FASTQ-to-VCF variant calling use alterlab-nf-core-sarek. Part of the AlterLab Academic Skills suite.
Scaffolds TÜBİTAK ARDEB proposals in the directorate's own section order: 1001 against the official .doc form (TR/EN özet ≤600 words each, ÖZGÜN DEĞER, YÖNTEM, PROJE YÖNETİMİ with iş paketleri and B Planı, YAYGIN ETKİ, EK-1/EK-2; ≤25 pages), and the PBS-typed 1002-A Hızlı Destek and 3501 Kariyer Geliştirme forms with per-section word ranges. Checks program caps (1001 ≤36 months, 3,000,000 TRY; 1002-A ≤12 months, 150,000 TRY/year; 3501 ≤36 months, 1,500,000 TRY, PI ≤7 years post-PhD, doçent or below), drafts to each program's criteria (1001 panel 35/25/20/20; 3501 adds kariyer geliştirme potansiyeli) and applies TÜBİTAK's generative-AI (ÜYZ) disclosure rules; submission via PBS after ARBİS. Delegates generic grant craft to alterlab-research-grants and BİDEB fellowships to alterlab-tubitak-bideb. Use when the user wants to write a TÜBİTAK 1001, 1002-A or 3501 proposal, draft özgün değer or yaygın etki sections, or map broader-impacts framing to TÜBİTAK terms. Part of the AlterLab Academic Skills suite.
Analyze an unfamiliar repository and explain what it does, how it runs, what architectural choices define it, where the important code lives, and what deserves deeper inspection next. Use this whenever a user has just cloned a repo, wants onboarding help, asks for a repo walkthrough, or needs a reliable first-pass architecture analysis.
Build an animation from scratch, making the decisions in the order that determines whether it feels right — should it animate at all, what purpose, which tool, which properties, which curve and duration, how it interrupts, how it exits. Writes the implementation. Use when asked to animate something, add motion, make a component feel alive, or build a transition. For critiquing existing motion use review-animations; for auditing a whole codebase use improve-animations.
Provides comprehensive guidance for Ant Design React component library including components, design system, themes, and TypeScript support. Use when the user asks about Ant Design for React, needs to build React applications with Ant Design, or implement design system patterns.
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.