Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Productivity
Marginal Revolution from your terminal via the public RSS feed. Trigger phrases: `check marginal revolution`, `search marginal revolution`, `latest MR posts`, `marginal revolution links`, `use marginalrevolution`, `run marginalrevolution`.
Plan and review products end to end across strategy, market research, product design, UI, monetization, distribution, operations, promotion, customer support, and lifecycle management. Use when designing a new app, SaaS, game, utility, web product, desktop product, product roadmap, launch plan, or full product operating system.
Use DeepSeek’s official Search / Fetch API to supplement search research, cross-check facts, find the latest information and sources, and fetch webpage Markdown content. Use when the user says “search with DeepSeek,” “DS search,” “supplement the search,” “find a few more sources,” “fetch this webpage with DeepSeek,” or configures a qiaomu-ds-search API Key. Search with sources and fetch public pages. If the API Key is missing, guide the user to the official page to create a key, configure it securely, and test it. Excludes offline-only rewriting or translation, local file search, chat model calls, general configuration that doesn’t specify DeepSeek, bypassing login walls, and site-wide crawling; supplements existing search capabilities rather than forcibly taking over all searches.
When the user is influenced by group pressure or mainstream opinion in decision-making and needs to distinguish "conformity = safety" from "independent analysis = truly conservative." Language signals: "everyone says" / "all the analysts recommend it" / "this is industry practice" / "why not follow the crowd" / "isn't it safer to be conservative." Not applicable to: decisions already supported by independent analysis (when no correction is needed).
Automatically process multiple recruitment markdown documents, recognize images in the documents (including online image links), extract key information (position, salary, deadline, etc.), filter by user criteria, and generate a structured summary report
Evidence-first research workflow for external web/source research and internal codebase investigation. Use when the user invokes $research, asks to research, investigate, audit what exists, compare options, understand current implementation, gather evidence, inspect docs, look up current facts, or answer repo-grounded questions before implementation. Prefer optional parallel agents for broad, separable research lanes such as external sources, codebase evidence, docs, risks, and competitor or alternative analysis.
Plan rebuttals, reviewer responses, and revision strategies for research papers. Use this skill whenever the user shares peer reviews, reviewer concerns, meta-review text, acceptance/rejection feedback, or asks how to respond to Reviewer 2, what experiments to run for rebuttal, what to concede, how to phrase a response, or how to prioritize limited rebuttal space. This skill focuses on strategy and response structure; use results-writeup for writing new result prose after the strategy is decided.
Use when starting work in a terminal session—opening a project or directory, connecting to a session by name, cloning a repo to work in, or running a command or coding agent somewhere other than the current shell. Sesh resolves names and directories to sessions across tmux, zoxide, config, and tmuxinator. Use instead of `tmux new-session`, `tmux new-window`, `tmux attach`, or `tmux send-keys`.
Search and monitor social media using X/Twitter (via Grok) and Reddit APIs. USE FOR: - Searching X/Twitter posts by keywords or hashtags - Finding X/Twitter users by criteria - Getting a user's recent posts - Searching Reddit posts and discussions - Getting comments from Reddit threads - Social media monitoring and research TRIGGERS: - "twitter", "X", "tweets", "posts on X" - "reddit", "subreddit", "reddit discussion" - "what are people saying", "social media", "sentiment" - "trending", "viral", "popular posts" - "user's posts", "timeline", "recent activity" Use x402.fetch for Grok (X) and Reddit endpoints. All endpoints are $0.02 per call. IMPORTANT: Never guess endpoint paths. All paths follow the pattern https://enrichx402.com/api/{provider}/{action}. Use exact URLs from the Quick Reference table below or call x402.discover_api_endpoints first.
Prospective failure analysis using Gary Klein's pre-mortem technique. Assumes complete failure, works backward to identify risks, leading indicators, and circuit breakers. Counters optimism bias by forcing systematic exploration of failure modes before they materialize. Use for project plans, architecture decisions, technology adoption, business strategy, or feature launches. Triggers on "리스크", "위험", "실패하면", "swing-mortem", "뭐가 잘못될 수 있어", "risk", "what could go wrong", "걱정되는 점", "failure modes", "리스크 분석", "위험 분석".
Generate probability-weighted alternative options that challenge default thinking. Forces unconventional alternatives and exposes hidden assumptions behind the "obvious" choice. For decision-point analysis, NOT full design exploration. Triggers on "대안", "alternatives", "옵션 뽑아", "options", "어떤 방법이", "아이디어", "다른 방법", "선택지".
Unified content extraction and action planning. Use when user says "tapestry <URL>", "weave <URL>", "help me plan <URL>", "extract and plan <URL>", "make this actionable <URL>", or similar phrases indicating they want to extract content and create an action plan. Automatically detects content type (YouTube video, article, PDF) and processes accordingly.
Review one work item for semantic quality and release-milestone compatibility, standalone or as the shared assessment used by task preparation and triage. Use when a work item's semantic quality and release-milestone compatibility must be assessed before publication. Russian discovery terms: проверить задачу.
Operate TaskContext projects and issues through authenticated SQL reads and REST writes. Use for triage, assignment, dependencies, comments, completion evidence, and issue history. Needs TASKCONTEXT_URL and TASKCONTEXT_USER_EMAIL, plus Google OAuth login or TASKCONTEXT_USER_PASSWORD.
Read Telegram channels and groups for financial news and market research using tdl (read-only). Use this skill whenever the user wants to read Telegram channels, export messages from financial Telegram groups, list their Telegram chats, search for news in Telegram channels, or gather market intelligence from Telegram. Triggers include: "check my Telegram", "read Telegram channel", "Telegram news", "what's new in my Telegram channels", "export messages from", "list my Telegram chats", "financial news on Telegram", "crypto Telegram", "market news Telegram", any mention of Telegram in context of reading financial news, crypto signals, or market research. This skill is READ-ONLY — it does NOT support sending messages, joining channels, or any write operations.
Extract explicit and hidden assumptions from a proof, model, or empirical argument. Use for theorem reading, paper review, proposal critique, debugging proofs, conclusions that appear stronger than the stated setup, or determining which premises are logically necessary, proof conveniences, and where the argument breaks.
Smart desktop and directory organizer. Automatically classifies and organizes files and application icons based on user prompts, removes duplicate versions, and generates an organization report. Supports organizing the desktop or a specified directory.
Personal travel concierge — complete itinerary planning and guide export from origin to destination. Input the origin, destination, number of days, budget, and style preferences to automatically output a closed-loop itinerary including transport recommendations, hotel recommendations, food routes, and a daily budget, with an optional generated guide. Use when the user mentions "做攻略", "旅行规划", "旅游计划", or "行程安排".
Turn a travel itinerary into a polished, offline-readable, mobile-first single-file HTML (interactive map + daily timeline + pre-departure ticket reminders). Two ways to use it — give it only a destination and number of days and let it plan for you, or hand it an existing plan and let it produce the page directly. Triggers: 旅行计划可视化、做旅行攻略网页、行程 HTML、travel plan visualization.
Generate comprehensive knowledge graph reports for Obsidian vaults. Combines multiple graph queries (vault-stats, hubs, bridges, orphans-rich) to produce a structured 4-module analysis report with data summary tables. Requires obsidian-graph-query skill installed. Use for: vault report, knowledge report, vault analysis, vault health, knowledge graph report, 知識圖譜報告, vault 分析.
Select and prioritize publication venues (conferences, workshops, journals) for a research paper. Use whenever the user asks where to submit, which conference fits best, whether a paper is ready for a top venue, how to resubmit after rejection, or when comparing venues like NeurIPS vs. ICLR vs. ICML vs. ACL vs. EMNLP vs. CVPR vs. AAAI vs. ICDM vs. journals. Also use when the user has a contribution in hand and needs to decide between tier levels, theory vs. empirical tracks, workshop vs. main track, or short vs. long paper format. If the paper isn't ready for any venue yet, use submission-readiness-audit first.
A complete tool for searching and reading WeChat official account articles, supporting keyword search and full-text extraction. **You must use this skill whenever the user mentions any of the following scenarios:** (1) searching official account articles or finding WeChat content by keyword (2) reading, summarizing, analyzing, or translating content from an mp.weixin.qq.com link (3) the user says "help me find official account articles", "read this WeChat article", "search official accounts", "scrape official accounts" (4) the user pastes a WeChat article URL and asks to extract or process its content (5) you need to get news, reports, or industry information from the WeChat ecosystem. Even if the user does not explicitly say "WeChat", if it involves retrieving or extracting official account content, this skill should be triggered. Triggers: 搜索公众号, 微信文章, 读这篇公众号, 抓取公众号, wechat article, mp.weixin.qq.com, 公众号内容, 微信链接
X Intelligence CLI for searching, monitoring, and analyzing X/Twitter from the terminal. Use for X/Twitter research prompts such as "search x", "find tweets about", "what's X saying", trend checks, follower diffs, real-time watch tasks, sentiment/report analysis, and bookmark-sync workflows. Best when recent community discourse matters (library/API changes, launches, market or cultural events). Supports search, watch, trends, bookmarks, likes/following (OAuth), Grok analysis, and JSON/JSONL/CSV/Markdown export. Non-goals: posting original tweets, DMs, or enterprise-only features.
Convert YouTube transcripts into summaries, notes, raw exports, or structured dossiers with timestamps. Use for YouTube dossier, video to notes, transcript to summary, or summarize YouTube video. Do NOT use for non-YouTube video, audio transcription, or downloading video files.
Extract and risk-rate hidden assumptions in a product brief or PRD. Use when asked to review a product brief for assumptions, audit a PRD for risks, find hidden assumptions, validate product plans, or run an assumption analysis. Produces a prioritised assumption map with confidence and impact scores, recommended validation methods, and critical assumption flags.
Transform saved links, papers, articles, posts, videos, and reference collections into approachable AI teaching artifacts for later study. Use when a user wants to queue learning material, create a readable explanation from a source, teach a paper or post step by step, or run an interactive tutor that validates understanding over multiple sessions.
Automates macOS Calendar via JXA with AppleScript dictionary discovery. Use when asked to "create calendar events", "automate calendar", "JXA calendar scripting", "EventKit automation", or "PyXA calendar automation". Covers events, calendars, recurrence, time zones, batch operations, and EventKit ObjC bridge.
Map publications, prior successful cases, datasets, benchmarks, and method families for banking and fintech AI problems. Use when the user needs publication-grounded evidence for credit risk, fraud, AML, collections, customer intelligence, personalization, banking document AI, compliance, contact-center automation, or financial-services R&D proposals. This skill specializes literature-triangulation for banking domain transfer and should feed publication-grounded-solution-design.
Subscribe to YouTube / Bilibili / podcast channels in BibiGPT, pull the latest feed, and mark items seen. Use when the user wants channel subscriptions, “what's new in my feed,” a daily digest, or to subscribe / unsubscribe. Do not use to summarize a new URL (`bibi`), search already saved videos (`bibi-library`), or analyze frames / mind maps (`bibi-vision`). Triggers: “subscribe to this channel,” “what's new in my subscriptions,” “daily digest,” “mark feed seen,” “订阅频道,” “我的订阅,” “有什么更新,” “取消订阅,” “latest feed.” Same `bibi` CLI binary and same BibiGPT account as the other skills.
Browse, search, and manage a BibiGPT saved library: already-summarized videos, personal notes, and collections. Use when the user asks what they already saved, to search past summaries, edit a note, or organize videos into a collection. Do not use to summarize a new URL (that is the `bibi` skill), follow channels / pull a subscription feed (`bibi-feed`), or analyze frames / generate a mind map (`bibi-vision`). Triggers: “what have I summarized,” “search my library,” “my saved videos,” “my notes,” “create a collection,” “add to collection,” “我之前总结过的,” “资料库,” “搜我的笔记,” “合集,” “列出最近的总结.” Same `bibi` CLI binary and same BibiGPT account as the other skills.
Scrape BOSS直聘 (job listing site) via Chrome CDP. Searches jobs by keyword/city/filters, fetches JD details, outputs structured JSON/CSV with plaintext salary, and can summarize scraped results into a job-market prompt. Optional --company locally filters results by company name (boss_name substring) with no extra request. Use when user wants to search/analyze jobs on BOSS直聘 or zhipin.com.
Projects 30-90 days of daily cash flow per cash account by combining current balance, scheduled recurring inflows and outflows from recurring.json, and a 6-month rolling average of discretionary spend, then flags days where projected balance falls below a configurable safety floor. Use for forward planning, detecting upcoming overdrafts, sizing pre-funding for sinking funds, or when user mentions cash flow projection, runway, balance forecast, or upcoming bills coverage.
Activate when the user is considering making a decision in an unfamiliar field, or needs to judge whether they truly "understand" something. Typical trigger signals: the user says "I can learn it", "it should be about right", or "this isn't hard", but lacks deep experience in the relevant field. Not applicable to scenarios where the user already has a long track record of success in the field and is acting within their circle, nor to early exploration stages where taking risks and trial and error are needed to expand boundaries.
Design growth models with compounding effects. Use when you have been doing something for a long time but growth is slow and you want to give up, or when you want your efforts to produce an "the more you do, the faster it goes" effect. Core method: find the pivot (verify causality) → design a reinforcing loop (A reinforces B, and B in turn reinforces A) → persist until the milestone. Not applicable to one-off tasks, short-term projects, or events with unclear causality. Key trigger signals: "why is there still no effect after doing it so long", "should I keep going", "how to make growth accelerate".
Compile timestamped conversation logs, agent sessions, IM messages, collaborative documents, meeting notes, task records, and similar OpenViking materials into concise, evidence-grounded daily reports. Use with ov compile when the user asks for a daily work report, end-of-day digest, 日报, or reports for one or more dates, especially when activities must be reconstructed across heterogeneous sources without treating plans or agent suggestions as completed work.
Use when a risk-and-reward decision of "take it / take it after negotiation / pause / decline" is needed for outsourcing, projects, or requirements — especially when scope/acceptance/payment/compliance/timeline/dependencies are unclear, or when high-uncertainty signals appear such as emergency queue-jumping, renewal changes, multiple-requirement prioritization, or AI/LLM work.
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.