Skills
100,928 packaged capabilities and playbooks for agents. Install any of them with one command.
All
Generate AI-implementable product requirement documents (PRDs) from one-line product ideas, vague app/site/tool requests, or feature concepts. Use when the user asks for a PRD, product requirements document, requirements document, AI-executable PRD, product specifications, an MVP plan, app/site/tool planning, or developer handoff docs for AI coding assistants
Records an architecture decision as a Nygard/MADR-shaped ADR under decisions/ — capturing the context that forced the choice, the options weighed, the decision itself, and its consequences in both directions, plus the supersedes chain that keeps a decision log honest. Read when asked to record an architecture decision, write an ADR, log the decision we made, document why we chose X over Y, capture this decision for the record, or supersede an old decision with a new one. Do NOT read to frame a proposal or explore an idea not yet decided (frame-a-proposal), to write a spec or implementation plan (write-a-spec), to write an incident postmortem (write-a-postmortem), or to judge whether a design is sound (review-a-design). This skill records a decision already made; it does not make one.
Reviews whether a design is SOUND — solving the right problem, derived from its stated goals and constraints — and emits ranked, evidence-backed findings, not edits. Read when asked to “review this design”, “is this design sound”, “pressure-test this proposal”, “do a design review”, “does this solve the right problem”, “poke holes in this spec”, “should we build this”, or to critique a proposal / spec / ADR / architecture or product decision. Do NOT read when the user wants to AUTHOR one of these — routing a new proposal is frame-a-proposal, a spec is write-a-spec, a decision record is record-a-decision, a postmortem is write-a-postmortem. Do NOT read for code review of a diff, or to fact-check individual claims (that is a correctness pass, a different job)
Generates Robot Framework tests in keyword-driven syntax with Python. Supports SeleniumLibrary, RequestsLibrary, and custom keywords. Use when the user mentions "Robot Framework", "*** Test Cases ***", "SeleniumLibrary", ".robot file". Triggers on: "Robot Framework", "*** Test Cases ***", ".robot", "SeleniumLibrary", "keyword-driven test".
Read an academic paper end to end and extract professional research insights, figures, metadata, and critique. Use this skill whenever the user shares a scientific paper, review paper, survey paper, systematic review, meta-analysis, scoping review, arXiv link, DOI, PDF, or pasted paper text and asks to read, summarize, analyze, extract, digest, review, critique, or explain it. For original research papers, produce a modified Heilmeier analysis. For review literature, produce a field-map extraction covering scope, taxonomy, evidence quality, consensus, controversies, gaps, and future directions. Do NOT use this skill for non-academic articles, blog posts, or news
Specialized workflows for drafting, refining, and responding to academic literature reviews and peer review feedback. Use this skill for literature review outlines, research-gap synthesis, reviewer rebuttals, response letters, and academic writing tone repair.
Send and receive faxes programmatically with Sinch Fax API. Use when building fax workflows, fax-to-email delivery, sending PDFs by fax, checking fax status, managing fax services, configuring cover pages, receiving fax webhooks, or integrating fax into healthcare, legal, or financial applications.
Interview the user in depth about what they actually want to build, then scaffold a working full-stack web app around it. Use when the user wants to start a new app, website, prototype, or SaaS; when they don't know what tech stack to pick; or when they want a solid working starting point fast. Covers requirements discovery, project setup, database (SQLite or Postgres in Docker), sign-in, transactional email, file uploads, payments, AI features, background jobs, optional agent access over MCP so tools like Claude can use the app, a real landing page and dashboard, optional public help documentation, a design system agreed with the user and written to DESIGN.md so every page matches, an account settings area with system logs and debugging built in, the legal pages and cookie consent an app of that kind actually owes its users, whether search engines and AI crawlers should find it at all, and a closing pass that proves the result actually builds, serves and does what was agreed rather than taking the builder's word for it
Transform podcast transcripts into multiple content assets—blog posts, social snippets, newsletters, and SEO-optimized landing pages—using systematic repurposing workflows. Use when: maximizing ROI from podcast episodes; creating blog content from audio/video; generating social media posts from long-form content; building newsletter content from transcripts; extracting quotes and highlights for promotion
Remove AI slop from research writing so papers, theses, grant proposals, reviews and rebuttals read as written by a careful human expert. Covers stock vocabulary (delve, tapestry, pivotal, underscores), empty emphasis, hedge stacks, formulaic signposting, "not only X but also Y" constructions, em-dash pileups, monotone rhythm, and claims vaguer than the data. Use when drafting or revising academic text with an AI assistant, when a draft "sounds like ChatGPT", before submission, or to match an author's own voice. Includes a zero-dependency slop linter for Markdown, LaTeX, text and Word files.
Search the web and read page contents without API keys. Use when you need to search via DuckDuckGo/Brave/Google (multi-page), extract readable text from URLs, browse interactively with a persistent visible browser (with tabs, click, screenshot, text search), download files/PDFs, or dismiss cookie banners. Supports JSON/markdown/text output. Powered by Playwright + Chromium.
Wren Engine CLI workflow guide for AI agents. Answer data questions end-to-end using the wren CLI: gather schema context, recall past queries, write SQL through the MDL semantic layer, execute, and learn from confirmed results. Use when: user asks a data question, requests a report or analysis, asks about metrics, revenue, customers, orders, trends, or any business data; user says 'how many', 'show me', 'what is the', 'top N', 'compare', 'trend', 'growth', 'breakdown'; user wants to explore, analyze, filter, aggregate, or summarize data from a database; agent needs to query data, connect a data source, handle errors, or manage MDL changes via the wren CLI.
Turn articles, scripts, or technical topics into 16:9 handwritten HTML slides in Wang Hong’s academic presentation style: near-white paper, small dark-blue handwriting, short red lines, thin red frames, green highlighter strokes, red warning triangles, word mind maps, and annotated screenshots. Produce an offline-ready single-file full-page version, a step-by-step writing animation version, per-slide PNGs, and an overview image (optional PDF) in one go. Works on macOS / Windows / Linux. Triggers: 王虹PPT风格、王虹手写PPT、Notability academic handwritten slides, handwritten web PPT, handwritten PPT, mathematician handwritten report style, handwritten PPT timeline version, handwritten presentation animation.
Expert knowledge for Azure Container Registry development, including troubleshooting, best practices, decision-making, architecture and design patterns, limits and quotas, security, configuration, integrations and coding patterns, and deployment. Use when configuring ACR Tasks, geo-replication, connected registries, image signing, or Kubernetes/ACI integrations, and for other Azure Container Registry development tasks. Not for Azure Container Apps (use azure-container-apps), Azure Container Instances (use azure-container-instances), Azure Kubernetes Service (AKS) (use azure-kubernetes-service), or Azure Red Hat OpenShift (use azure-redhat-openshift)
Expert knowledge for Azure Data Factory development including troubleshooting, best practices, decision making, architecture & design patterns, limits & quotas, security, configuration, integrations & coding patterns, and deployment. Use when configuring IR/SSIS/SHIR, mapping data flows, SAP/SaaS connectors, CI/CD deployments, or secure VNets/Key Vault, and other Azure Data Factory related development tasks. Not for Azure Data Explorer (use azure-data-explorer), Azure Synapse Analytics (use azure-synapse-analytics), Azure Stream Analytics (use azure-stream-analytics), Azure Databricks (use azure-databricks)
Expert knowledge for Azure Speech in Foundry Tools development including troubleshooting, best practices, decision making, limits & quotas, security, configuration, integrations & coding patterns, and deployment. Use when using STT/TTS containers, SSML/avatars, Voice Live agents, telephony APIs, or MAI-Transcribe workloads, and other Azure Speech in Foundry Tools related development tasks. Not for Azure Content Understanding in Foundry Tools (use azure-content-understanding), Azure AI Language (use azure-language-service), Azure Translator (use azure-translator), Azure AI Bot Service (use azure-bot-service).
Master Gary Halbert's direct-response copywriting principles from “The Boron Letters” (1984). These timeless fundamentals separate pros from amateurs. Use for writing direct-response copy (sales letters, emails, ads); creating attention-grabbing headlines; building and segmenting mailing lists; improving copy through the A-pile test; or learning copywriting fundamentals from scratch.
Claude Code + Codex parallel pipeline for bootstrapping Trellis coding specs. CC analyzes the repo with GitNexus (knowledge graph) + ABCoder (AST), creates Trellis task PRDs with full architectural context and MCP tool instructions, then Codex agents run those tasks in parallel to fill spec files. Use when: bootstrapping coding guidelines, setting up Trellis specs, 'bootstrap specs for codex', 'create spec tasks', 'CC + Codex spec pipeline', 'initialize coding guidelines with code intelligence'. Also triggers when user wants to set up GitNexus or ABCoder MCP for multi-agent spec generation.
Multi-baseline counterfactual cost analysis. Compares actual session spend to hypothetical always-haiku / always-sonnet / always-opus routing baselines. Answers "is the routing earning its keep?" Negative savings flag over-escalation; positive savings quantify the router's win.
Forward-looking spend extrapolation. Computes a USD-per-day rate from the recent measurement window, projects to 7d/30d/90d/365d horizons, and surfaces "days until budget exhausted" when a budget is configured. Predictive counterpart to `cost-budget-check` (reactive).
Per-message cost breakdown within a single session. The drill-down companion to cost-anomaly — when an outlier session is flagged, this surfaces the specific expensive messages so operators can see whether the cost came from output tokens, cache writes, or model escalations.
A cram engine for finals—an AI teacher grounded in the science of learning. Based on cognitive load theory, elaboration, the generation effect, and retrieval practice, its four-stage pipeline—break down → teach → quiz → fill gaps—turns course highlights into interactive, efficient study sessions. Launch with one command and save students from finals.
Generate a project-local skill that formally verifies the code where a bug costs the most and tests reach the least, in any language, with whichever verification tool fits. Use for /create-formal-verification-skill, “model-check this race”, “prove this invariant”, or when concurrency, retries, state machines, parsers, or unsafe code need more than tests.
Image prompt templates, model selection guidance, and anti-generic patterns for generating visual assets. Use when the user needs AI-generated images for landing pages, marketing, or products. Covers hero images, feature illustrations, OG cards, icons, and backgrounds
Manage DatoCMS records, uploads, and project automation through the Content Management API (CMA). Use for content reads/edits, publish/unpublish, bulk imports/exports, CSV scripts, localization, Structured Text and block edits, assets, roles/tokens, webhooks, scheduling, environment operations, and typed Node.js/TypeScript CMA scripts. Includes short follow-ups such as “publish them” or “fix those slugs”, token/OAuth setup before content operations, and model/field changes needing a migration-versus-direct decision or embedded in CMA automation. Works with the CLI or an available DatoCMS MCP; neither MCP installation nor local execution is assumed. Route schema-inspection-only requests, versioned migrations, and standalone CLI configuration to `datocms-cli`, GraphQL reads to `datocms-cda`, and frontend or plugin code to their dedicated skills.
Generates Detox E2E tests for React Native apps in JavaScript. A gray-box testing framework with automatic synchronization. Supports local simulators/emulators and the TestMu AI cloud. Use when the user mentions “Detox”, “React Native test”, “element(by.id())”, or “device.launchApp”. Triggers on: “Detox”, “React Native E2E”, “React Native test”, “element(by.id)”, “device.launchApp”.
Generate visual diagrams in standard Draw.io (.drawio) format for deep learning models, network architectures, algorithm workflows, and more. Supports both creation from scratch and style transfer. From scratch: model architecture diagrams, flowcharts, receptive field illustrations, etc. Style transfer: reference image + content description/project → generate a new diagram in the reference image's style. Ensures the XML format is valid and can be opened and edited directly in Draw.io.
Environment-specific adaptation rules for EdgeOne Makers Skills running in sandboxed or restricted AI coding environments (e.g. WorkBuddy). Trigger when the user is working in WorkBuddy or a sandboxed IDE where CLI prompts hang and the network is proxy-isolated (but browser login still works). Covers non-interactive CLI flags, network isolation workarounds, login in a sandbox, proxy bypass, file preview constraints (MUST use http:// via a dev server, NEVER file://, NEVER python -m http.server / npx serve), and dev server requirements.
A brand product-launch email — masthead with wordmark, hero image block, headline lockup with skewed-italic accent, body copy, primary CTA, and a specifications grid. Pure HTML email layout (centered single column, table fallback). Use when the brief asks for an "email", "newsletter blast", "MJML", "product launch email", or "email template".
Build RAG pipelines with Exa.ai for real-time web retrieval. Use when building retrieval-augmented generation, integrating Exa with LangChain, LlamaIndex, Vercel AI SDK, or implementing AI agents with web search capabilities. Triggers on: RAG pipeline, retrieval augmented generation, Exa LangChain, Exa LlamaIndex, ExaSearchRetriever, ExaSearchResults, Exa MCP, Exa tool calling, Claude tool use, AI agent web search, grounded generation, citation generation, fact checking, hallucination detection, OpenAI compatibility, chat completions.
Exa.ai search API integration for neural and keyword web search with content retrieval. Use when implementing web search features, integrating Exa SDK (exa_py, exa-js), or retrieving web content. Triggers on: Exa, exa_py, exa-js, neural search, web search API, search_and_contents, searchAndContents, find_similar, findSimilar, domain filtering, date filtering, text extraction, page summaries, highlights, search auto mode, fast search, search categories, livecrawl, excluding domains, include text, exclude text, EXA_API_KEY.
Plan and run an early-stage fundraising process and produce a Fundraising Pack (raise decision memo, round design brief, pitch narrative + deck outline, investor pipeline + tracker, outreach/follow-up scripts, diligence checklist). Use for fundraising, raising capital, venture capital, pitch deck, investor outreach, pre-seed, seed. Category: Career.
Deploy Python applications to Google App Engine Standard/Flexible. Covers app.yaml configuration, Cloud SQL socket connections, Cloud Storage for static files, scaling settings, and environment variables. Use when: deploying to App Engine, configuring app.yaml, connecting Cloud SQL, setting up static file serving, or troubleshooting 502 errors, cold starts, or memory limits.
Not legal advice. Generate Hebrew documents (PDF, DOCX/Word, PPTX) with correct RTL layout, mixed Hebrew-and-English bidi handling, and Hebrew typography. Use whenever the output is a Hebrew or mixed Hebrew/English Word document, Hebrew PDF, or Hebrew PowerPoint ("Hebrew Word document", "מסמך Word בעברית", "create a .docx in Hebrew"), or Israeli templates like Heshbonit Mas, Hozeh, or Protokol. ALSO use this for the symptom where a Hebrew document looks fine on screen or in Claude but comes out scrambled, reversed, or broken in Word, with English, numbers, or punctuation on the wrong side ("Hebrew text reversed in Word", "fix Hebrew formatting in Word"); the fix is regenerating the .docx with paragraph-level RTL/bidi, NOT a web/CSS RTL change. Prefer over the generic docx/pdf skills ONLY when the document is Hebrew or RTL; for English-only docs use the generic skill. Covers reportlab, WeasyPrint, python-docx, pptxgenjs. Do NOT use for OCR or reading existing documents (use hebrew-ocr-forms).
Expert-level Helm 3 package management, chart development, templating, and production operations. Use when the user mentions Kubernetes, package manager, charts, templates, or deployment, or when the task involves Helm Architecture, Chart Structure, Values.yaml, or Helm Hooks.
Execute Hugging Face Hub operations using the `hf` CLI. Use when the user needs to download models/datasets/spaces, upload files to Hub repositories, create repos, manage local cache, or run compute jobs on HF infrastructure. Covers authentication, file transfers, repository creation, cache operations, and cloud compute
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.