Skills
96,770 packaged capabilities and playbooks for agents. Install any of them with one command.
Productivity
Use when the user wants to generate and literature-vet a pool of novel, testable research hypotheses for a question or domain. A multi-agent loop: a Generator proposes candidate hypotheses, a LiteratureScout grounds each in real retrieved literature (already known? closest prior work? what gap does it fill?), and a Judge scores them against a fixed rubric and keeps the strong, non-duplicate ones; rounds repeat — mutating toward the open gaps — until fresh rounds stop adding keepers. Not for sharpening or decomposing a research question (no grounding/scoring there), and not for grading an existing written proposal against the literature.
Create or expand an Idea.md / IDEA.md file from a rough description, existing repo, conversation history, notes, or other early-stage product inputs. Use when the user asks to "write an Idea.md", "turn this into an idea file", "capture this product idea", "expand this concept", or wants a repo-grounded concept brief before validation, PRD, or implementation work.
AI mock interviewer for big tech. Covers top internet/tech companies at home and abroad such as Alibaba, Tencent, ByteDance, Baidu, Meituan, JD, Huawei, Didi, Pinduoduo, Google, Meta, Amazon and Microsoft. The user enters the target company name, role name and JD content, uploads a resume, and the agent plays the big-tech interviewer, combining the JD requirements with the resume to output 10 high-quality interview questions (with difficulty, reference answer hints and follow-up directions), and supports generating "good answer vs bad answer" comparison examples, dedicated HR interview training, salary negotiation phrasing guidance, and a coherent multi-round full interview simulation. Trigger words: big tech interview, overseas big tech interview, mock interview, interviewer, help me interview, interview preparation, interview practice, mock interview, I want to interview at ByteDance, Alibaba interview questions, Google interview, Meta interview, Amazon interview, AWS interview, Microsoft interview, help me come up with 10 interview questions, good answer vs bad answer, sample answer, answer comparison, mock HR interview, reason for leaving, salary negotiation, career planning, offer negotiation, salary phrasing, full interview process, multi-round simulation, first/second/third round.
Jacky Shen's Agile Coach AI assistant, providing structured prompt support for Scrum Masters and Product Owners. This skill MUST be activated whenever the user mentions any of the following Scrum, Sprint, Backlog, user stories, DoD, DoR Definition of Ready, Product Owner, Scrum Master, agile ceremonies, retrospectives, backlog refinement, sprint planning, release planning, risk identification, user personas, customer churn, product requirements, decision-making, scenario modeling, team facilitation, or any agile team workflow. Keywords are Scrum, Sprint, Backlog, user story, Product Owner, Scrum Master, DoR, agile, iteration, retrospective, prioritization, risk, persona, churn, PRD, facilitation, minimal experiment, root cause, pattern recognition. Whenever a user mentions agile team work, PO/SM scenarios, or needs structured agile prompt templates, this skill must be used.
Drive a desktop application from one plain-language goal, or resolve one intent at a time, without ever reading the accessibility tree. Use when automating a desktop app and you want to keep a 150-element JSON tree out of your context. Give `run.mjs` a whole goal and it observes, decides, and acts until the goal is met; give `act.mjs` a single step when you want to keep the plan yourself. Covers which element, which operation, when to look deeper, and when to stop. Requires a clone of the agent-desktop repository and a TYPESAFE_API_KEY.
Linear ticket lifecycle for any project. Use when starting work on a Linear issue, ending work on an issue, or asking what to work on next. Triggers include, start KAT-N, pick up, implement, finish, complete, done with, close, what's next, next ticket, next issue. Handles blocker validation, status transitions, context loading, branch creation, evidence gating, and chain promotion.
Save links, articles, bookmarks, YouTube videos with transcripts, PDFs, audio, and notes as clean Markdown, then search and read them later. Keep is a read-it-later app, bookmark manager, and web clipper for AI agents. Use it when the user says "save this link", "clip this page", "save for later", "remember this", or "keep this", and search the user's saved library before asking them to repeat something they read, watched, or wrote. Keep syncs X (Twitter) bookmarks and imports from Pocket, Instapaper, Raindrop, Omnivore, and Kindle highlights. Create or update Markdown notes for ideas, decisions, and research. Use the Keep CLI or the Keep MCP server, whichever is available.
Researches public LinkedIn company, product, and showcase pages via the Crawlora API. Look up known IDs or discover and search the public product directory by keyword or category. Does not access personal profiles or private content.
Fill web forms by fetching form fields from a URL, deep-searching the user's local knowledge base for relevant info, and generating a markdown document with all answers pre-filled. Use when the user provides a URL to a web form (conference application, speaker submission, event registration, profile form) and wants help filling it out from their existing materials. Also trigger when the user mentions "填网页表", "fill web form", "网页填表", "表单填写", "申请表填写", "conference application", "speaker submission", "讲师申请", "报名表", or provides a URL with "form", "feedback", "apply", "register", "submit" in the path.
Register a new matter — a unified questionnaire covering identifying information, conflict checks, source, risk triage, materiality, outside counsel, internal owner, evidence preservation, and key dates; writes matter.md and history.md and appends a structured line to _log.yaml. Use when the user says "new matter", "register this matter", or needs to add a new matter to the portfolio.
Apply a latticework of mental models from multiple disciplines to improve decision quality. Use this skill when the user needs to think more clearly, avoid cognitive blind spots, apply cross-disciplinary reasoning, or evaluate a complex decision from multiple angles — even if they say 'how should I think about this', 'what am I missing', 'give me a different perspective', or 'what frameworks apply here'.
Conduct scenario planning to prepare for multiple plausible futures using driving forces, uncertainty axes, and the 2x2 scenario matrix. Use this skill when the user faces high uncertainty, needs to stress-test a strategy against different futures, or prepare contingency plans — even if they say 'what if things go wrong', 'what could the future look like', 'how do we prepare for uncertainty', or 'stress-test our strategy'.
Retrospective on a money-making opportunity in progress: record actual time invested/actual income vs the original expectations, capture lessons learned (which kinds of opportunities work for me, which pitfalls I have hit), and update the status file. Trigger words: "复盘"/"这个搞得怎么样"/"实际赚了多少"/"money retro"/"总结一下".
View your personal task history and status on OpenAnt. Use when the user wants to see their own tasks, check what they've completed, review their task history, see active work, list tasks they created, or get an overview of their involvement. Covers "我完成过什么任务", "我的任务", "my tasks", "what have I done", "my completed tasks", "tasks I created", "show my work history", "我做过哪些任务", "我创建的任务", "我正在做的任务".
Implicit use requires an explicit or already-established Nature Portfolio context. Design multi-source scholarly searches, literature maps, search strategies, citation-network or independent-citation audits, deduplicated evidence tables, and field-level verification of existing bibliographies. Within that scope, use for 文献检索、综述检索、MeSH、查论文、核对参考文献, systematic discovery, ref checks, or bibliography cleanup. Do not use to attach sources to drafted claims (nature-citation), download full text, or run alerts.
Use this to pull a specific figure, fact, quote, or table value out of a collection of documents and cite the exact source file and page. Built for question-answering over a folder of reports — annual reports, 10-Ks and financial filings, research PDFs, scanned forms / images (`.jpg` `.png` `.tiff`), Office (`.docx` `.pptx`), HTML / TXT, audio, and video. It indexes the whole corpus once, then finds the right document among many and returns the value with its page number — useful when several documents or figures look alike and you need the correct one, not a near-match. Reach for it instead of reading or grepping PDFs one at a time. Not for: editing files, web browsing, a single plain-text file, fine-tuning.
Search, read, and write Notion pages, databases, and blocks. Use this whenever the user wants to find a page in Notion, read a database, add a row, create or append content to a page, or asks "what's in my Notion" — even if they don't say "API". Also use it for any URL under notion.so or a mention of a Notion page/database ID. Always start from this skill when interacting with this service — its bundled scripts and recipes are the fastest path.
Use when the user gives a journal or magazine name and wants to know submission requirements, submission format/style checklists, official journal website information, community reviews, review speed, fees, article types, format details, or asks "how do I submit to this journal / is it reputable / research the submission guide". Verifies the official journal website and community reviews online, and by default isolates all intermediate files to `.bensz-api/task-{yyyymmdd-hhmm}-{short-description}/paper-know-journal/` in the current working directory, delivering only `KnowJournal-{journal-name}.md` at the end. ⚠️ Not applicable: when the user wants to select a journal for a manuscript (use paper-select-journal), write the paper body, download full-text PDFs, or just ask a simple fact that needs no finished document.
Maintain a personal CRM — track relationships, last-touch dates, follow-ups owed, context from past conversations, and who deserves a check-in. Use this whenever the user says 'who do I owe a follow-up to,' 'when did I last talk to,' 'personal CRM,' 'track this contact,' 'log this conversation,' 'who should I check in with,' 'add [name] to my contacts,' or asks about their relationships and follow-up history. The skill maintains a structured `contacts/` folder with one markdown file per person, automatically surfaces follow-ups, and pulls context before any meeting.
Generate rich personality profiles from social media data exports (Twitter/X, LinkedIn, Instagram). Use when a user wants to analyze their social media presence, create a personality profile for AI personalization, understand their communication patterns, or extract insights from their digital footprint. Triggers on requests like "analyze my Twitter data", "create a personality profile", "what can you learn about me from my posts", "personalize an AI for me", or when users provide social media export files.
Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.
Run a Pragmatic Institute Gap Analysis — a market-driven self-assessment of how well the team performs the 37 activities of the Pragmatic Framework, who owns each, and where the biggest gaps are. Use when the PM asks “where do we stand”, “how are we doing on the framework”, a product-org self-assessment, a maturity check, or where to focus. Produces a Gap Analysis table plus a top-3-gaps action plan as a Markdown or Word .docx artifact.
Build a Pragmatic Institute Strategy Matrix — a forced-choice 4×4 that decides who to win and what to build over a set time frame, and surfaces alignment between the product team and leadership. Use when the PM asks about product strategy, “where should we invest”, “what should we build for whom”, focus, prioritizing markets, or avoiding a spread-too-thin roadmap. Produces a Strategy Matrix as a Markdown or Word .docx artifact.
Write market-driven, prioritized product requirements (a Market Requirements Document / MRD) the Pragmatic Institute way. Use when the PM asks for “requirements”, “an MRD”, “what should we build”, “prioritize features”, “market requirements”, or wants development to stop saying “requirements aren't specific enough”. Articulates each requirement as a persona's problem, prioritizes with Market Evidence × Impact, and groups them for delivery. Produces a Market Requirements Document as a Markdown or Word .docx artifact.
Debug failing Power Automate cloud flows using the FlowStudio MCP server. The Graph API only shows top-level status codes. This skill gives your agent action-level inputs and outputs to find the actual root cause. Load this skill when asked to: debug a flow, investigate a failed run, why is this flow failing, inspect action outputs, find the root cause of a flow error, fix a broken Power Automate flow, diagnose a timeout, trace a DynamicOperationRequestFailure, check connector auth errors, read error details from a run, or troubleshoot expression failures. Requires a FlowStudio MCP subscription — see https://mcp.flowstudio.app
Foundation skill for Power Automate via FlowStudio MCP — auth setup, the reusable MCP helper (Python + Node.js), tool discovery via `list_skills` / `tool_search`, and oversized-response handling. Load this skill first when connecting an agent to Power Automate. For specialized workflows, load `power-automate-build`, `power-automate-debug`, `power-automate-monitoring` (Pro+), or `power-automate-governance` (Pro+) — each contains the workflow narrative, this skill provides the plumbing they all rely on. Requires a FlowStudio MCP subscription or compatible server — see https://mcp.flowstudio.app
Run a premortem on any plan, launch, product, hire, pricing change, strategy, or high-stakes decision. Uses Gary Klein's prospective-hindsight method — assume it already failed at a future date and work backward to surface every plan-specific cause, then produce a revised plan, mitigations, early-warning signals, and a pre-launch checklist. USE WHEN premortem this, run a premortem, what could kill this, future-proof this, stress test this plan, what am I missing here, find the blind spots, what could go wrong, am I missing anything, poke holes in this, where will this break, before I commit to this. NOT FOR adversarial attack on an arbitrary idea or argument (use RedTeam), current-state multi-perspective debate (use Council), network/system vulnerability testing (use Security), or simple feedback/factual questions. DO trigger when someone has a concrete plan or commitment where the cost of being wrong is high and the decision is still changeable.
Researches any project idea against live data from GitHub and Dev.to to surface what already exists, how mature the space is, and where the real opportunity lives. Use when a developer describes something they want to build and wants to know if it's been done before. Triggers on phrases like "validate my idea", "has this been built", "is this already a thing", "what exists for X", "should I build this", "is this idea original", "check if my project exists", "what are the alternatives to what I want to build", "is the market saturated for X", or any request to research the competitive landscape before starting a project.
Researches public Reddit posts, comments, communities, user history, and trends through the Crawlora API. It also supports explicitly requested buying-intent lead discovery that scores public posts. Use for community research, sentiment, trends, or a clearly requested lead search; do not infer sensitive traits or target people on that basis.
A research roadmap for legal questions — statutes to consult, case-law areas to investigate, the administrative regulatory framework, and search keywords for 北大法宝 / 法信 / 元典. Provides leads and a framework, not authoritative citations; students verify and develop everything. Use when a student asks where to start researching, needs a research roadmap for an issue, or needs to identify gaps in existing research.
Diagnose what's actually driving behavior — the person's own, someone else's, or a group's. Surface motivation vs. root motivation. Triggers: 'what motivates them', 'why are they doing this', 'what does this person actually want', 'how do I get buy-in', 'why won't they engage', 'what's really driving this', 'intrinsic vs extrinsic motivation', 'what need is this serving'.
Scan live job boards and salary databases to find real-time compensation data for any role and location. Use this skill when a user asks "what's the going rate for a senior React engineer in London", "software engineer salary Singapore", "how much do ML engineers make", "what should I be earning as a [role]", "is my salary competitive", "what does [company] pay for [role]", "salary range for [job title] in [city]", or any request to find out what a role pays in a specific market.
Alibaba Cloud Drive resource advisor: provides four kinds of workflows — inventory/organize/catalog/plan — to take stock of and organize resources, generate Obsidian collections and incremental Excel/family navigation, and generate study plans based on the learning situation the user provides. Triggers: "organize cloud drive", "cloud drive inventory", "update cloud drive index", "update Excel master index", "generate family navigation Excel", "build a library for the cloud drive", "make a study plan from cloud drive resources".
Pause and critically reflect on the direction of the current conversation. Use this whenever the user says 'take a step back,' 'let's zoom out,' 'are we sure about this,' 'wait — am I missing something,' 'sanity check this,' 'pressure test this,' 'are we going down the wrong path,' or expresses doubt about where the discussion has gone. Also use proactively if you notice the conversation has accumulated assumptions that haven't been examined, or has drifted from the user's original goal. Output is a structured reflection — original goal, path taken, assumptions made, what might be wrong, and a recommendation.
Submit completed work for a task on OpenAnt. Submission = text description + files. IMPORTANT — before submitting, always check if your work produced any files and upload them first. Use when the agent has finished work and wants to deliver results, submit a solution, turn in deliverables, upload files, or send proof of completion. Covers "submit work/task", "deliver results", "I'm done", "here's my work", "submit solution", "upload and submit", "attach proof", "deliver file", "send deliverable".
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.