Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Productivity
Compare intended product outcomes against observed results to close the launch-to-learning loop: collect post-launch evidence, distinguish expected from observed from uncertain from inferred claims, update assumptions, assess feature health, and choose among continue/improve/harvest/pivot/pause/retire — including retirement lifecycles with deprecation, migration, customer treatment, and retained reusable learning. Do not use for incident postmortems or root-cause analysis (routes to incident-learning or site-reliability-engineering); do not use for analytics instrumentation or metric dashboard design (routes to product-analytics-and-measurement); do not use arbitrary thresholds as universal retirement rules — decisions require human judgment and context.
Design industry R&D solution directions grounded in prior successful cases, academic papers, technical reports, and credible industry publications. Use whenever the user has a vague business, banking, fintech, enterprise, or operational problem and wants to brainstorm methods, draft a theoretically motivated solution, compare solution families, or justify a proposal with publications instead of unsupported intuition. Also use when the user asks for "prior successful cases", "paper-backed solutions", "method draft", "research-backed proposal", or "what has worked before" for an applied industry problem.
Stress-test a research idea before major investment. Use to expose hidden assumptions, weak novelty, confounders, and the cheapest decisive experiment. Use before starting a project, writing a proposal, scaling experiments, or committing major compute/resources.
Gate on identity questions to the user. Before any "who is X?" (or role / relationship question) reaches the user, exhaust the brain's lookup chain: think → search + page read → mounted sources → timeline/graph → web. If escalation survives the chain, ask WITH a hypothesis, never a bare unknown. Also owns the no-placeholders-at-ingest rule: pages created during bulk ingestion get their relationship/role resolved immediately.
Schedules sales meetings efficiently — booking pages, round-robin routing, calendar integration, reminders, no-show recovery, and meeting page conversion. Use when prospects aren't booking meetings, no-show rates are too high, round-robin isn't distributing evenly, booking page conversion is low, unsure which scheduling tool to use, or Mixmax meetings aren't working. For Yesware-specific help, use /sales-yesware. Do NOT use for lead routing rules (use /sales-lead-routing), cadence design (use /sales-cadence), or calendar-based CRM automation (use /sales-integration).
Personal knowledge base powered by Ensue for capturing and retrieving understanding. Use when user wants to save knowledge, recall what they know, manage their toolbox, or build on past learnings. Triggers on "save this", "remember", "what do I know about", "add to toolbox", "my notes on", "store this concept".
Operate Slack workspaces from a terminal or agent: list channels, read messages, follow threads, search message history, list files, and verify inbound webhook signatures — with a bundled slack-cli script that is read-only by default and gates every send behind a --dry-run/--yes confirmation. Use when an agent needs to read or post Slack data, triage incidents, or answer questions about what was said in a workspace. Do not use for building Slack apps or bots (that is application development) or workspace administration like user provisioning and org settings (that is the Slack admin console).
SupSub WeChat official account discovery — use only when the user wants to find / discover a WeChat official account itself (usually to get its mpId and then subscribe to it). Matches "搜公众号 X", "找 X 这个公众号", "查 X 公众号", "想订阅 X 公众号但不知道 ID", "discover / find a WeChat MP account". ⚠️ Do not use to search the body/content of articles or articles by keyword — that intent (e.g. "搜一下 X 相关的内容", "找 X 相关的文章") must go through supsub-search. Also covers cancelling a running official account search task.
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.
Manage technical projects from initiation through delivery and closure: choose and tailor predictive, iterative, Scrum, flow-based, or hybrid approaches; coordinate teams and vendors; track milestones, risks, issues, dependencies, forecasts, scope changes, and recovery. Use when a team has no project manager, a TPM needs decision-ready analysis, a project is slipping, or commitments and handoffs need coordination. Use `technical-program-management` instead when multiple related projects, teams, vendors, shared capacity, or benefits must be coordinated toward one outcome. Adapt depth to the user's expertise. Do not use for a standalone coding task, product discovery or strategy, only writing an implementation plan, only diagnosing Kanban flow, operating a ticket tool, incident command, or a release go/no-go decision; route those to specialists.
Pressure-test mathematical arguments, derivations, and ML paper claims, and deliver a written report-style audit. Use for theorem sketches, proof outlines, step-by-step derivations, ML paper claims, reviewer-mode critique, hidden assumptions, skipped proof steps, invalid generalization, weak empirical support, overclaimed conclusions, or whenever a result looks cleaner or stronger than expected — including "check my derivation", "which step breaks", "did I drop a constant", "does this algebra hold", or "audit this proof". Use argument-audit instead when the user wants to be walked Socratically to the weak link rather than handed a report, or when the claim is an essay/application/everyday argument rather than a formal or empirical one.
Use when asked to “thinking in bets,” “make decisions under uncertainty,” “think probabilistically,” “avoid resulting,” “separate decision quality from outcomes,” or “reduce bias in decisions.” Helps make explicit bets and evaluate decisions on process, not results. The Thinking in Bets framework (from Annie Duke) applies poker strategy to business and life decisions.
Use when the user wants web research: gathering cited, multi-angle evidence on a specific question. Triggers on: "research X for me", "do web research on", "look up sources for", "find citations for", "gather evidence on", "what does the web say about X". Also invoked programmatically by other beagle skills (prfaq-beagle Ignition, brainstorm-beagle reference points, strategy-interview context grounding) via the companion contract. Does NOT trigger on codebase lookups ("find this function", "search the repo"), local file search, LLM-as-judge evaluation, or paywalled/auth-gated scraping. Produces a written plan, parallel-subagent findings, and a cited synthesis report on disk — never inline prose, never unsourced claims.
Edit AI-sounding text so it reads like a careful person wrote it, without changing what it says or commits to. Built for work writing where the wording matters: sales and marketing copy, support replies, HR and legal letters, emails, docs, release notes, posts, landing pages. Two modes. Rewrite: edit the text you are given. Interview: when a draft is AI-composed and thin, ask the writer 3-5 sharp questions and build the piece from their answers, in their words. Also writes from the writer's own notes or a voice-memo transcript. Calibrated on measured differences between human and AI writing.
Route Zephyr entries from any document to their proper bake logs, with deduplication, reverse chronological ordering, and move or copy semantics. Use when asked to route Zephyr entries, move entries out of a log to where they belong, or clean up a document holding mixed entries.
Web search for current or external info. Trigger: search verbs ("search", "查", "搜", "google"), freshness ("latest", "最新", "news", "更新", "进展"), comparison ("调研", "对比", "vs"), proper nouns. Skip: local ops, explicit other search tool, general knowledge.
Harvests article metadata, abstracts, and full-text PDFs from DergiPark (TÜBİTAK ULAKBİM's national journal-hosting platform, 3,000+ journals) via its verified platform-wide OAI-PMH endpoint (https://dergipark.org.tr/api/public/oai/; verbs Identify/ListSets/ListRecords/GetRecord; prefixes oai_dc/oai_mods/oai_marc/oai_etdms; setSpec=journal-slug), parses Highwire citation_* and DC.* meta tags on /pub/{slug}/article/{id} pages, pulls PDFs from the citation_pdf_url path, and emits BibTeX/RIS locally. Use when the user wants to harvest a Turkish journal, fetch DergiPark articles, list a journal archive, get BibTeX/RIS for a DergiPark paper, or read a journal aim-and-scope (öz/kapsam). For TR Dizin indexing status use alterlab-trdizin; for YÖK theses use alterlab-yok-tez; for academic profiles use alterlab-yok-akademik. Part of the AlterLab Academic Skills suite.
Andrej Karpathy's thinking frameworks and way of expressing ideas. Systematically distilled from 20+ blog posts, 16 in-depth interviews and 100+ X posts, extracting 6 core mental models, 8 decision heuristics, complete Chinese output adaptation and a quick reference of classic phrasings. Use as a thinking advisor, analyzing the reliability of AI technology, learning methods, industry trends and product design from Karpathy's perspective. Activates when the user mentions "from Karpathy's perspective", "what would Karpathy think", "Karpathy" or "karpathy mode". Also applies to: Software 2.0/3.0 discussions, vibe coding topics, neural network training, AI hype assessment and the limits of LLM capabilities. It can also trigger when the user merely says "from an engineering realism angle", "march of nines", "building is understanding" or "jagged intelligence". Do not trigger when the user is simply asking ordinary AI-related questions — activate only when they explicitly want Karpathy-style thinking frameworks.
Read an academic-career applicant profile from files, folders, notes, CVs, resumes, transcripts, research statements, project writeups, publication lists, GitHub summaries, recommendation notes, and target-school materials. Use this skill whenever the user asks Claude to read "my profile", "my application folder", "my CV and notes", "my PhD materials", "my academic background", or wants a reusable structured profile before SOP writing, academic CV building, professor emails, faculty/program fit mapping, recommender packets, or application-package audits. This skill is for academic career and research-application contexts; do not use it for generic job-search resumes, industry recruiter profiles, literature surveys, or writing a final SOP/CV directly.
Decision-quality pre-commitment analysis for initiative-level go/no-go calls with meaningful downside and limited reversibility — launches, rewrites, key hires, partnerships, strategic bets, timelines. Invoke when the user explicitly asks to pressure test, validate, or evaluate such a decision (e.g. 'should we do this,' 'pressure test,' 'go/no-go,' 'are we ready'). Do NOT invoke for routine code review, sub-task planning, reversible tactical choices, or pure ideation.
Given a research proposal, claim, method, or result, recommend the baselines and benchmarks needed to support it — split into a MINIMAL set (necessary for credibility, skipping any risks rejection) and a SUGGESTED set (complementary, strengthens external validity). Use whenever the user proposes something and asks "what should I compare against", "what baselines do I need", "which benchmarks support this claim", "is this enough to be convincing", "what's the minimal experiment set", "what would a reviewer demand", or shares a method/idea and wants the comparison plan. Default to running this on any proposal so claims are backed by the right evidence. Always check first whether the input is sufficient (precise claim, field, the novel component, metric); if not, request the missing context before recommending rather than guessing. Use experiment-design to design the run itself (or to design a measurement when no benchmark exists), and hypothesis-and-ablation-planner for internal ablations.
Systematically review biblical content through a 10-point quality checklist. Works with bible-buddy reference files, any file path, URLs, or text pasted directly in the conversation. Use when asked to review, audit, check, or verify biblical content. Trigger on: "檢查冷知識", "fact check", "檢查經文", "review bible", "幫我檢查這段", "bible-fact-check", or any request to find errors, duplicates, or issues in biblical reference content.
Creates and configures monday.com boards with proper structure including groups, columns, and initial items. monday.com boards are flexible enough to manage any workflow - project trackers, CRM pipelines, bug trackers, content calendars, and more. Use when user asks to "create a board", "set up a project", "new board for", "build a tracker", "organize my project", "create a board for tracking", or "set up a workflow".
Business rules elicitation and analysis techniques. Covers rule types (constraints, derivations, inferences), decision tables, rule templates, and policy documentation. Use when identifying business policies, constraints, calculations, and decision logic during requirements elicitation.
Use when verifying bibliography citations in a document or when the user pastes academic citations to check, especially when worried about LLM hallucinations where real DOIs or NBER working paper numbers are paired with fabricated titles. Queries Crossref canonical metadata and flags metadata mismatches.
This skill should be used when the user is building, planning, or strategizing and the key question is whether to optimize content (what) or change form (how/medium). Trigger on "내용 vs 형식", "content vs form", "metamedium", "형식을 바꿔볼까", "새로운 포맷", "관점 전환", "perspective shift", "다른 방법 없을까", "같은 방식이 안 먹혀", "diminishing returns". Applies Alan Kay's metamedium concept to surface form-level alternatives. For requirement clarification use vague; for strategy blind spots use unknown.
This skill should be used when the user provides a strategy, plan, or decision document and wants to surface hidden assumptions and blind spots using the Known/Unknown 4-quadrant framework. Trigger on "known unknown", "4분면 분석", "blind spots", "뭘 놓치고 있지", "뭘 모르는지 모르겠어", "전략 점검", "전략 분석", "assumption check", "가정 점검", "quadrant analysis", "what am I missing". Strategy-level blind spot analysis with hypothesis-driven questioning. For requirement clarification use vague; for content-vs-form reframing use metamedium.
Daily news briefing aggregator that fetches and delivers content from 5 sources — Readhub (Chinese tech news), Karpathy-recommended RSS feeds (92 HN popular blogs), GitHub Trending, Hacker News top stories, and V2EX hot topics — with summaries for each item. Use when setting up a daily automated briefing, running an on-demand news digest, or configuring a morning push notification via cron. Triggers on "每日资讯", "daily briefing", "早报", "新闻推送", or requests to aggregate news from HN/GitHub/V2EX/Readhub.
Extract language learning content from web pages and create structured flashcards in markdown format for spaced repetition study (Anki decks). Works with ANY language (Spanish, Japanese, Arabic, Chinese, etc.). Use when the user asks to create flashcards, flash cards, study cards, Anki decks, or mentions spaced repetition from online content. Supports grammar lessons, vocabulary, and exercises. Converts to Anki-importable format.
Guide a person in cultivating creativity in their own work and life: open conversational sessions on creative blocks, habits, environment, motivation, and resilience, or structured development of a concrete project or fledgling idea through a five-phase practice. Do not use for therapy or clinical support, general life coaching, product or stakeholder discovery, or as a study guide for a book.
Lark video meetings: query meeting records and artifacts (minutes/transcripts/Miaoji), search/upload/download/edit Miaoji, and have bots join meetings; query ongoing meetings and real-time meeting content (speaking/chat/shared documents), answer questions about meetings, and send in-meeting chat messages/reactions. Look up related information using meeting_id, meeting_no, event_id, note_id, minute_token, vc-node-id, or a Miaoji URL. Use lark-calendar for scheduling, availability, and room management.
Activate when the user analyzes a complex problem from only a single perspective (e.g. purely financial / purely technical / purely market), or lacks an analytical framework for a cross-domain problem. Typical triggers: the user says "I only understand XX" or "this problem seems to be an XX problem" while clearly relying on a single discipline's explanation. Not applicable when the user already has a mature multidimensional analytical framework and is using it effectively, nor to simple purely execution-level tasks.
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.