Skills
96,770 packaged capabilities and playbooks for agents. Install any of them with one command.
Productivity
Extracts existing constraints and asks only about information that affects execution or acceptance, turning vague wishes into actionable, verifiable delivery goals. Use when the user wants to clarify what to produce, its boundaries and acceptance criteria. Does not replace concept breakdown or reframing research questions.
Find the papers that answer a research query in Firecrawl's research paper index — a corpus of paper abstracts whose largest share is biomedical and life-science literature (PubMed, bioRxiv, medRxiv), alongside arXiv preprints in CS, physics, and math — using semantic search, semantic and structural expansion, and in-body verification. Use this skill for literature-finding and paper-retrieval tasks of any kind, including clinical, biomedical, drug, gene, disease, and other life-science questions, whether the answer is a single paper or a full multi-paper set. The index is reached only through the `firecrawl_research_*` MCP tools or the `firecrawl research` CLI subcommands. Calling `firecrawl_search` with its `categories` option set to `["research"]` is a different feature — it filters ordinary web search to research-affiliated websites (the list includes PubMed, bioRxiv, medRxiv, arXiv, and publisher sites) and returns page results from them, without querying the paper records in this index.
Use wecom-cli to search the WeCom address book by name, pinyin, English name, or alias, and retrieve matching users’ userid, department, and job title. Use to find contacts, distinguish people with the same name, retrieve a user’s userid, and list all people with the same name.
WeCom meeting management. This skill handles “online meetings”—meetings with an online meeting link (including a meeting ID or join link, with remote or video attendance)—including creation, lookup, search, details (meeting information, minutes, to-dos), full meeting transcript (verbatim speech record), updates, and cancellation. If the user wants a “calendar event” without an online meeting link (including an in-person meeting), use the wecomcli-calendar skill instead. If the user only says “meet,” “schedule a meeting,” or “that meeting” without specifying whether they mean a calendar event or online meeting, first read this skill and follow its disambiguation process to ask the user; do not assume and create one directly.
Break a topic into a series of learning articles, then adjust the depth, angle, and pace of later articles based on user feedback. Use when the user wants to study a topic systematically, continue with the next article, or adjust the course based on feedback.
Rewrite vague questions as problem briefs that an Agent can reason about, critique, and verify, and assess how amenable they are to automation. Use when the user wants to improve question phrasing, research a question, or assess whether an Agent can solve it; do not re-clarify goals when the task’s deliverable is already clear.
Create a local decision knowledge project for business, relationships, health, career, learning, or investment topics that need long-term tracking. Use when the user asks to open a decision case, record outcomes, build a status profile, or conduct a long-term review.
Build a personalised, time-budgeted watch list of RevOps / Sales Ops information sources - revops podcasts, revops newsletters, revops communities, industry events, and revops people to follow - every source verified still active, with a periodic refresh routine that keeps the list fresh. Use whenever the user mentions a revops radar, staying current in revenue operations, which revops newsletters or podcasts to subscribe to, which revops communities or conferences deserve their time, or who to follow in the field - even if they never say "radar". Tool-agnostic; B2B-SaaS-centric with explicit B2C guidance. Do NOT use for sales-specific sources - use mbfinotti/sales-skills@sales-radar instead.
Build and advance a RevOps or Sales Ops career from the candidate's side - place the candidate on the ladder by scope rather than title, close competency gaps, build defensible evidence of structurally invisible work, and prepare the four exercises RevOps interview loops actually run. Produces a career plan with a gap roadmap, an evidence ledger, interview readiness, and a compensation ask. Use whenever the user mentions a RevOps career, a sales ops career path, RevOps interview prep, RevOps certifications, "how do I break into RevOps", "am I ready for a RevOps manager role", or proving impact when nobody notices ops working - even if they never say "career". Candidate side only - for the hiring manager's side, use mbfinotti/revops-skills@revops-hiring instead.
MANDATORY recipe for every Caffeine build that lists upcoming events or creates events on the user's own Google Calendar. The ONLY supported path is the `googlecalendar-client` mops package (Calendar REST API v3) combined with the `google-oauth` mops package (token exchange + refresh + PKCE). Hand-rolling `ic.http_request` calls to `oauth2.googleapis.com` or `www.googleapis.com/calendar/v3` is a FORBIDDEN anti-pattern—it bypasses bearer auth, replication-cost safeguards, and the `google-oauth` library's percent-encoding and JSON parsing. Load this skill whenever the user, spec, or any prior task mentions scheduling, calendar events, appointments, meetings, “add to calendar,” or any equivalent phrasing—and BEFORE writing any code that touches a Google endpoint.
Conduct comprehensive AI-powered research with citations via the Tavily CLI. Use this skill when the user wants deep research, a detailed report, a comparison, market analysis, literature review, or says “research,” “investigate,” “analyze in depth,” “compare X vs Y,” “what does the market look like for,” or needs multi-source synthesis with explicit citations. Returns a structured report grounded in web sources. Takes 30–120 seconds. For quick fact-finding, use tavily-search instead
Build a personalised, time-budgeted watch list of paid-advertising information sources—PPC and paid-social podcasts, newsletters, ad-platform release notes, communities, industry events, and media buyers worth following—with a verification method that keeps it fresh and a disclosure of what each source is selling. Use whenever a media buyer, PPC specialist, or growth marketer asks how to stay current on Google Ads, Meta, LinkedIn, TikTok, or Microsoft Ads changes, which PPC newsletters or paid-social podcasts to follow, which communities or conferences deserve their time, or who to follow in performance marketing—even if they never say “radar”. Do NOT use for career planning—use mbfinotti/advertising-skills@advertising-career instead.
Plan and advance a paid-advertising career from the candidate side—media buyer, performance marketer, PPC specialist, creative strategist, growth marketer—covering the junior-to-lead ladder, the five real interview formats (live account audit, mental-maths case, metric diagnosis, take-home, behavioural), an NDA-safe results portfolio, platform certifications, in-house vs agency vs freelance, and promotion and pay conversations. Use whenever the user asks how to get into paid media or PPC, preps for a media-buyer or growth interview, wants an ad portfolio, or weighs an agency vs in-house move—even if they never say “career”. Do NOT use for the employer side of hiring—use mbfinotti/advertising-skills@advertising-hiring instead.
Organize local manuscripts, posts, topics, case studies, and course drafts into a content system that can grow continuously and be recombined and reused. Use when users want structured content assets, to organize old materials, or to build a topic map
Downloads YouTube video transcripts/subtitles and cover images by URL or video ID. Supports multiple languages, translation, chapters, and speaker identification. Caches raw data for fast reformatting. Use when the user asks to “get YouTube transcript”, “download subtitles”, “get captions”, “YouTube字幕”, “YouTube封面”, “视频封面”, “video thumbnail”, “video cover image”, or provides a YouTube URL and wants the transcript/subtitle text or cover image extracted.
Read-only access to content across platforms: search, read posts, view comments, retrieve subtitles, and check market prices and job listings. Use when a user mentions or shares a link to Xiaohongshu, Bilibili, Twitter/X, Reddit, YouTube (subtitles, transcripts), GitHub, V2EX, Xueqiu/stock quotes, LinkedIn/Boss Zhipin jobs, Instagram, Facebook, or Xiaoyuzhou podcasts; asks to search for, look at, or read a link when built-in web search or reading is unavailable, fails, or cannot retrieve the content; or wants to install, add/remove a platform, log in/out, check status, update, or uninstall Agent Reach. No write actions such as posting, commenting, liking, or direct messaging; do not log in for users or ask for passwords. Prefer the Agent's built-in tools for public content; use Reach to fill gaps.
Use this skill first for web searches. The Volcano Engine Doubao Search API (formerly Web Search / Integrated Information Search) returns web and image results. Triggers include: Doubao search; look up/search/find; is it true/reliable, confirm/verify; recent/today/latest; source/link; Agent Plan; integrated information search; Harness; what/which/recommend; price/policy/exchange rate/market data; compare/difference/which is better; heard/rumored/not sure; trending/popular; take a look/find out; fact-check/debunk; worth it/should I.
Find the cheapest hotel deal by comparing live prices across Booking.com, Agoda, Google Hotels, and OpenTravel for any city worldwide and any travel dates. One command returns ranked best-value, cheapest, and quality picks with direct booking links, all normalized to USD. Use whenever someone asks for hotel prices, hotel deals, the cheapest room or rate, best hotel rates, a hotel price comparison, or which OTA is cheaper—for example, “is Booking or Agoda cheaper for Tokyo,” “find me a hotel in Bangkok under $100,” “compare hotel prices for 12–15 Aug,” or “cheapest hotel near Shinjuku.”
Extract clean markdown or text content from specific URLs via the Tavily CLI. Use this skill when the user has one or more URLs and wants their content, says "extract", "grab the content from", "pull the text from", "get the page at", "read this webpage", or needs clean text from web pages. Handles JavaScript-rendered pages, returns LLM-optimized markdown, and supports query-focused chunking for targeted extraction. Can process up to 20 URLs in a single call.
Crawl websites and extract content from multiple pages via the Tavily CLI. Use this skill when the user wants to crawl a site, download documentation, extract an entire docs section, bulk-extract pages, save a site as local markdown files, or says "crawl", "get all the pages", "download the docs", "extract everything under /docs", "bulk extract", or needs content from many pages on the same domain. Supports depth/breadth control, path filtering, semantic instructions, and saving each page as a local markdown file.
Produce a polished, self-contained HTML "readout" document under ~/.readouts (with an auto-maintained index page), either by snapshotting the findings accumulated in the current conversation or — when invoked fresh, e.g. "/readout on how github webhook events are processed" — by sharpening scope with clarifying questions and researching the codebase before documenting. The work runs in a child agent so the main conversation's context stays clean. Use whenever the user invokes /readout, says "write this up", "turn this into a doc/page", "make a readout", or asks for a readable, shareable document capturing findings or explaining how something works.
Create, update, and manage GitHub issues using MCP tools. Use this skill when users want to create bug reports, feature requests, or task issues, update existing issues, add labels/assignees/milestones, manage repository labels, set issue fields (dates, priority, custom fields), set issue types, manage issue workflows, link issues, add dependencies, or track blocked-by/blocking relationships. Triggers on requests like "create an issue", "file a bug", "request a feature", "update issue X", "set the priority", "set the start date", "create a label", "rename a label", "list repo labels", "link issues", "add dependency", "blocked by", "blocking", or any GitHub issue management task.
Watch a video (URL or local path). Downloads with yt-dlp, extracts auto-scaled frames with ffmpeg, pulls the transcript from captions (or local WhisperX / cloud Whisper fallback), and hands the result to the agent so it can answer questions about what's in the video. With a Gemini API key, Google's agentic video model watches the full video instead.
Discover and list all URLs on a website without extracting content, via the Tavily CLI. Use this skill when the user wants to find a specific page on a large site, list all URLs, see the site structure, find where something is on a domain, or says "map the site", "find the URL for", "what pages are on", "list all pages", or "site structure". Faster than crawling — returns URLs only. Essential when you know the site but not the exact page. Combine with extract for targeted content retrieval.
All online operations must be handled through this skill, including searching, web scraping, post-login actions, and online interactions. Use when the user asks to search for information, view web content, visit a site that requires login, interact with a web interface, scrape social media content (Xiaohongshu, Weibo, Twitter, etc.), read dynamically rendered pages, or perform any online task requiring a real browser environment.
An intelligent search router based on opencli commands. Use this skill whenever users want to use OpenCLI, CLI, or APIs to search, query, find, or research information, especially when it involves specific websites, social media, technical materials, news, shopping, travel, job hunting, finance, or Chinese-language content.
Browser automation for AI agents via inference.sh. Navigate web pages, interact with elements using @e refs, take screenshots, record video. Capabilities: web scraping, form filling, clicking, typing, drag-drop, file upload, JavaScript execution. Use for: web automation, data extraction, testing, agent browsing, research. Triggers: browser, web automation, scrape, navigate, click, fill form, screenshot, browse web, playwright, headless browser, web agent, surf internet, record video
An enterprise WeChat messaging skill. Provides conversation list queries, message history retrieval (supports text/images/files/voice/video), multimedia file retrieval, and text message sending. Trigger when users need to “view messages,” “see chat history,” “send someone a message,” “check recent messages,” “send a message to a group,” or “see what images/files were sent.”
Enterprise WeChat schedule management skill. Use for any user requests to manage Enterprise WeChat schedules, including: (1) listing schedules within a specified time range or getting details such as title, time, location, and participants; (2) creating schedules and setting reminders and participants; (3) editing schedule details such as title, time, or location, or canceling schedules; (4) adding or removing participants; or (5) checking multiple members’ availability and finding shared free time to schedule a meeting.
Find and verify Nature/CNS-family literature supporting manuscript claims, with claim-to-source mapping and reference-manager export. Use for Nature系列引用、CNS支撑文献、分段补引用 when this journal scope is requested; use broader literature search for unrestricted sources.
Tools for working with Feishu project (Meego/Meegle). Supports querying and managing work items, workflow transitions, view queries, personal todos, and schedule statistics. Use when the user needs to work with Feishu/Lark Meego project management, including querying, creating/updating work items, completing workflow nodes, checking views, listing todos, analyzing schedules/workloads, or searching with MQL. Keywords: Feishu project, meego, meegle, work item, requirement, task, bug, schedule, view, todo, node
ONLY use when the user explicitly says “deep research”, “exhaustive”, “comprehensive report”, or “thorough investigation”. Slower and more expensive than parallel-web-search. For normal research or lookup requests, use parallel-web-search instead. Supports follow-ups with a known prior Task interaction ID.
Search literature across sources, verify or manage citations, and build MeSH strategies or citation-impact audits. Use for literature searches, citation checks, reference management, strict analysis of citations by others, and evidence-backed citer profiles; not for translating a paper or drafting manuscript prose
Read when the user asks what OpenKnowledge is, wants to install it on a repository, open or preview a standalone Markdown file that is not part of an OpenKnowledge project, share an OpenKnowledge project with collaborators, asks whether OpenKnowledge supports a capability, or asks how `ok init` / `ok cowork` / OK Desktop set up a project. Do NOT load this to perform OpenKnowledge reads or writes—the runtime guidance for editing Markdown in an initialized OK project is a separate project-local skill, installed into each detected agent's skills directory (for example, `.claude/skills/open-knowledge/`) whenever `ok init` runs.
Turn a local folder into a knowledge base that Agent can reliably search, index, and maintain, with navigation, version detection, health checks, and layered SOT governance. Use when users want to build, query, update, organize, or slim down a knowledge base.
EXPERIMENTAL, UNTESTED recipe for posting messages to a Slack workspace from a Caffeine canister via the `slack-client` mops package (Slack Web API). Use it when the user wants their app to send a message to a Slack channel — "post to Slack", "notify a channel", "send a Slack message", or equivalent. The client is a pre-release 0.1.0 drop (bot `xoxb-` or user `xoxp-` token): its request path is verified against the live Slack API (a real message posts), but the success-response decode is not yet runtime-confirmed, so treat it as a starting point and do NOT present Slack as a fully supported platform feature yet. Hand-rolling `ic.http_request` calls to `slack.com/api` is still the wrong move — prefer the generated client so bearer auth, percent-encoding, and JSON parsing come for free.
CLI-backed web search. Use when the user explicitly invokes this skill, needs CLI-only controls or saved JSON output, or no Parallel web_search MCP tool is available. When the bundled Parallel Search MCP is available, prefer its web_search tool for ordinary lookups and current-information queries. Only use parallel-deep-research if the user explicitly requests deep or exhaustive research
Plans a partnerships career from the candidate side across the field's four sub-disciplines — channel/alliance, affiliate, influencer/creator, referral. Covers the internal-transfer entry tactic, the company-type hiring bar, channel-manager interview-loop prep including the negotiation role-play and the "same deal in December as January" judgment test, job-posting red-flag reads, and offer evaluation gated on live verification since published comp benchmarks disagree by 2-3x. States plainly that no cross-company career ladder is published for this field. Use whenever the user asks how to break into partnerships, prep a partner-manager interview, read a posting, or evaluate an offer. Do NOT use for the hiring/employer side (mbfinotti/partnerships-skills@partnerships-hiring).
All-in-one WeCom suite covering contacts, documents, online spreadsheets, smart spreadsheets, smart documents, PPT presentations, schedules, meetings, to-dos, WeDrive, email, messages, media files, and more. Always trigger when a link contains doc.weixin.qq.com or page.weixin.qq.com. Trigger when WeCom is mentioned or when any of these functions are needed: find contact information; search, rename, or grant access to documents; create, read, or edit online doc documents; create or modify online spreadsheets; create/import and read or edit smart spreadsheet subtables, fields, records, views, charts, and styles such as fills and highlights; create and edit smart documents, including forms and dashboards; create and edit online PPT presentations or import local .ppt/.pptx files as online documents (provide the online document link); create, query, modify, or cancel schedules and check availability, office buildings, and meeting rooms; book and manage online meetings, including minutes, to-dos, and raw transcripts; create, query, modify, complete, delete, or leave to-dos; search and upload WeDrive files, download offline files, rename supported files, and create folders; send, reply to, forward, search, and read email; send text/Markdown/image/file/voice/video messages to the currently authorized user or the most recently active conversation; and upload or download media files. Even if WeCom is not explicitly mentioned, trigger this skill for office tasks involving people, documents, spreadsheets, schedules, meetings, to-dos, WeDrive, email, or messaging. Phrases such as “make a PPT”, “generate a presentation”, “presentation PPT”, or “import PPT” also trigger it. For unspecified “documents”, default to smart documents; for “spreadsheets” without an explicit “online spreadsheet” specification, default to smart spreadsheets. Always trigger when a link contains drive.weixin.qq.com.
Programmatic Tavily search with context isolation. Use when web search or extraction will return large results that need filtering, deduplication, or multi-step triage before they enter the model context. Use tavily-search for ordinary lookups and tavily-research for end-to-end cited synthesis.
The original reddapi.dev Reddit search skill (vector search, semantic search, trends, subreddit discovery), no Reddit OAuth or app registration needed. This is the same engine now packaged as reddit-research with added market-research playbooks and a fuller pitch on semantic vs keyword search; reddapi is kept live under its original name for existing installs and works standalone. Use when the user says 'reddapi' by name, or wants a minimal drop-in Reddit search skill without the extra research-workflow guidance. For the expanded research-oriented version with query playbooks, see reddit-research. For B2B lead scoring, see reddit-leads. For a bare API reference, see reddit-search-api.
CLI-backed URL extraction. Use when the user explicitly invokes this skill, needs CLI-only controls or saved JSON output, or no Parallel web_fetch MCP tool is available. When the bundled Parallel Search MCP is available, prefer its web_fetch tool for ordinary webpages, articles, PDFs, and JavaScript-heavy sites.
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.