sciver

132 items · Skills 33 · MCP 96 · Plugins 3

Research & knowledge

Answers with sources: web search, reading pages, documentation and project memory.

What you can make

  • research report with sources
  • competitor research
  • up-to-date API docs
  • knowledge base
Open in the library

With search, filters and ratings.

Ready-made packs

Ready-made tool sets for this task. Load any of them in your AI with one command.

  • Weekly report

    A report on the market and competitors from the past week, with sources – saved in Notion, with a summary for the team on Slack.

    /weekly-report

All items (132)

Skills (33)

  • Academic Research Skills

    Takes academic work from research through writing and review to the final version.

  • AI Research Skills

    98 skills for AI research: training, fine-tuning, interpretability and paper writing.

  • Brave Search Skills

    Skills for the Brave Search API: web, news, images, videos, places and AI answers.

  • Claude Scientific Writer

    Writes papers, literature reviews and grant proposals grounded in sources.

  • Deep Research (OpenAI)

    Refines a research question and runs deep research through the OpenAI API.

  • Deep Research Skills

    Runs deep research: a research plan, work by parallel subagents and a Markdown report.

  • Fact Checker

    Checks the facts in a document against official sources and proposes corrections.

  • Gemini Deep Research Skill

    Hands multi-step research to the Gemini Deep Research agent and returns a cited report.

  • Hugging Face Papers

    Reads Hugging Face Papers and arXiv papers along with their linked models and datasets.

  • last30days

    Researches any topic from the last 30 days across Reddit, X, YouTube, HN and Polymarket.

  • Litreview

    Plans a literature review, searches PubMed and OpenAlex and writes a DOCX guide.

  • Paper Preflight

    Checks every reference in a paper against real scholarly records before you submit.

  • Parallel Agent Skills

    Parallel skills for deep research, data enrichment, FindAll and web monitoring.

  • Scientific Agent Skills

    Teaches the agent scientific work: biology, chemistry, medicine and over 100 databases.

  • Tavily Skills

    Tavily CLI skills: search, research reports, company data and monitoring.

  • Apify Agent Skills

    Scrapes popular sites with ready-made Apify Actors and helps build your own Actors.

  • Azure Agent Skills (Microsoft Learn)

    200+ skills on Azure services, built from Microsoft Learn documentation.

  • BibiGPT Skill

    Summarizes YouTube, Bilibili and TikTok videos and podcasts, with chapters and transcripts.

  • Bright Data Skills

    Teaches the agent to collect web data through Bright Data and keep an eye on job costs.

  • Claudemy

    Turns any topic into a course of about 12 two-minute audio lessons with sources.

  • Designer Skills

    273 skills for designers: research, UX, design systems, UI, prototyping and critique.

  • Explorium GTM Skills

    Builds company and contact lists, enriches data, scores leads and writes cold emails.

  • Firecrawl Skills

    Searches, scrapes and crawls pages through the Firecrawl CLI and ready-made workflows.

  • Hugging Face Datasets

    Browses datasets on the Hugging Face Hub: splits, rows, search and filters.

  • Marker

    Converts PDFs, images and Office files to Markdown, JSON or HTML, with OCR.

  • MinerU

    Reads PDFs, scans and Office files with OCR, tables and formulas, citing pages and blocks.

  • Nexscope eCommerce Skills

    Skills for sellers on Amazon, Shopify, eBay, Etsy, TikTok Shop and Walmart.

  • Notion Research Documentation

    Finds answers across Notion sources and writes a cited brief, comparison or report.

  • Obsidian Second Brain

    Runs your Obsidian vault as a second brain: daily notes, people, decisions and kanban boards.

  • Obsidian Skills

    Teaches the agent Obsidian’s Markdown, Bases and Canvas formats and the Obsidian CLI.

  • SE Ranking SEO Skills

    26 SEO skills that turn SE Ranking data into content briefs, page audits and reports.

  • Weights & Biases Skills

    Logs and analyzes ML experiments in W&B, tunes hyperparameters and traces agents in Weave.

  • YouTube Skills (TranscriptAPI)

    Fetches YouTube transcripts, searches videos and browses channels and playlists.

MCP (96)

  • Context7

    Gives the agent up-to-date library docs instead of the ones from its training data.

  • Exa MCP

    Searches the web, reads pages as clean Markdown and runs multi-step research.

  • Firecrawl MCP

    Turns pages into Markdown or JSON, maps and crawls sites and searches the web.

  • Perplexity MCP

    Searches the web with filters, answers with sources and prepares in-depth reports.

  • Apify MCP

    Runs ready-made scrapers from Apify Store: maps, online stores, social media and more.

  • arXiv LaTeX MCP

    Fetches the LaTeX source of arXiv papers so the agent reads math formulas accurately.

  • arXiv MCP

    Finds and downloads arXiv papers, reads them by section, exports BibTeX and tracks topics.

  • Basic Memory

    Markdown notes as shared memory for you and the agent, with a knowledge graph and search.

  • BioMCP

    Searches 72 biomedical sources: PubMed, ClinVar, ClinicalTrials.gov, OncoKB and more.

  • Brave Search MCP

    Searches Brave for web pages, news, images, videos and places, and creates AI summaries.

  • Cognee MCP

    Long-term agent memory built on a knowledge graph: remember, recall and forget.

  • Consensus MCP

    Searches more than 400 million scholarly articles, including all of PubMed.

  • Deep Research MCP

    Delegates in-depth research to OpenAI, Gemini or Tavily and returns the finished report.

  • DeepWiki MCP

    Answers questions about public GitHub repositories based on their wikis.

  • DuckDuckGo Search MCP

    Searches the web through DuckDuckGo and fetches page content, with no account or API key.

  • Fetch MCP

    Fetches any web page and turns it into Markdown the agent can read.

  • Fetcher MCP

    Fetches pages through a Playwright browser, including JavaScript-heavy ones, as Markdown.

  • Graphiti MCP

    Builds a knowledge graph for the agent that tracks changes over time and searches it.

  • Hindsight

    Agent memory that learns from experience: it retains, recalls and reflects.

  • Hugging Face MCP

    Searches Hugging Face models, datasets, Spaces, documentation and papers.

  • Jina AI MCP

    Reads pages and PDFs as Markdown and searches the web and arXiv.

  • Kagi MCP

    Searches Kagi for web pages, news, videos and podcasts and fetches pages as Markdown.

  • Linkup MCP

    Searches the web with natural-language questions, fetches pages and runs deep research.

  • MCP Local RAG

    Searches your PDF, DOCX and Markdown documents locally without sending them to an API.

  • MCP Memory Service

    Persistent memory for agents: a knowledge graph, semantic search and a web dashboard.

  • MCP server for Obsidian

    Reads, searches and edits Obsidian notes through the Local REST API plugin.

  • MCPVault

    Reads, searches and edits Obsidian notes straight from the files, with no plugins.

  • MediaWiki MCP Server

    Searches and reads Wikipedia or any MediaWiki wiki and can edit pages once you sign in.

  • Memory MCP

    Remembers facts across sessions in a local graph of entities, relations and observations.

  • Microsoft Learn MCP

    Searches official Microsoft and Azure documentation and code samples.

  • Open-WebSearch

    Searches several search engines and fetches page content, with no API keys.

  • OpenZIM MCP

    Searches offline ZIM archives such as Wikipedia, Wiktionary or Stack Exchange.

  • Parallel MCP

    Searches the web, fetches pages as compact Markdown and delegates in-depth research.

  • PubMed MCP Server

    Searches PubMed and Europe PMC, fetches full text and citations and formats references.

  • Readwise MCP

    Searches your Readwise highlights and Reader documents and helps tidy your inbox.

  • Ref MCP

    Searches API and library documentation and returns only what is needed.

  • SearXNG MCP

    Private web search through your own SearXNG instance, with page reading.

  • Sequential Thinking MCP

    Gives the agent a tool for breaking a problem into steps, with revisions and branches.

  • SerpApi MCP

    Returns results from Google, Bing, YouTube and other search engines as data and reads pages.

  • Supermemory MCP

    Shared memory across AI apps: saves facts and recalls them in new conversations.

  • Tavily MCP

    Searches the web in real time, extracts page content, maps and crawls sites.

  • Valyu MCP

    Searches the web and specialized indexes: papers, SEC filings and clinical trials.

  • Wolfram MCP

    Computes exact answers and draws on the Wolfram|Alpha knowledgebase instead of guessing.

  • You.com MCP

    Searches the web and fetches page content through You.com; search alone works without a key.

  • Zotero MCP

    Searches, reads, cites and annotates papers from your Zotero library.

  • AgentQL MCP

    Extracts structured data from a page based on a plain-language description of the fields.

  • Ahrefs MCP

    Researches keywords, backlinks, traffic and competitors in Ahrefs and audits your site.

  • Android Source Explorer MCP

    Gives the agent AOSP and AndroidX source code so it doesn’t guess from stale knowledge.

  • anymd

    Turns PDFs, Office files, EPUB, web pages and scans into clean Markdown, locally.

  • Autodesk Product Help MCP

    Answers questions about AutoCAD, Revit, Fusion or Maya from Autodesk’s official help.

  • Bright Data MCP

    Fetches blocked pages, search engine results and data from more than 1,200 ready-made scrapers.

  • Browserbase MCP

    Controls a Browserbase cloud browser: runs plain-language actions and extracts data.

  • Crawl4AI MCP

    Scrapes and crawls pages into clean Markdown through the Crawl4AI cloud or your own server.

  • Dash MCP

    Searches documentation from the docsets installed in the Dash app on macOS.

  • Data Commons MCP

    Fetches public statistics, such as demographic and economic data, from the Data Commons graph.

  • DataForSEO MCP

    Pulls SERP results, keywords, backlinks and on-page audits from the DataForSEO API.

  • Decodo MCP

    Scrapes pages through Decodo proxies and pulls Google, Amazon, TikTok and YouTube results.

  • Design Systems MCP

    A knowledge base on design systems, WCAG and ARIA with dates and sources, available to the agent.

  • Discogs MCP Server

    Searches the Discogs database and manages your record collection, ratings and marketplace listings.

  • Docling MCP

    Converts PDFs into structured documents with layout and tables, and generates new ones.

  • Facebook Ads Library MCP

    Browses any brand’s ads in the Meta Ad Library and analyzes the creatives.

  • Fathom MCP

    Gives the agent access to meetings recorded by Fathom’s AI notetaker.

  • Fireflies MCP

    Searches Fireflies meeting transcripts, summaries and action items.

  • Google Developer Knowledge MCP

    Searches and reads up-to-date Google developer documentation, such as Google Cloud.

  • Granola MCP

    Searches Granola meeting notes and transcripts and organizes folders.

  • Grep MCP

    Searches code in about a million public GitHub repositories through grep.app.

  • Grounded Docs MCP

    Indexes library documentation for the versions your project uses and searches it locally.

  • LinkedIn MCP Server

    Reads LinkedIn profiles, companies, posts and your feed and sends messages after confirmation.

  • Listen Notes MCP

    Searches podcasts, episodes and transcripts and finds contacts for PR outreach.

  • LlamaParse MCP

    Parses documents to Markdown, classifies and splits them and extracts data against a schema.

  • Markdownify MCP

    Converts PDF, DOCX, XLSX, PPTX, images, audio and web pages to Markdown.

  • MarkItDown MCP

    Converts PDF, Word, Excel, PowerPoint, images and web pages to Markdown.

  • Microsoft Work IQ

    Asks Microsoft 365 Copilot about emails, meetings, documents and Teams chats.

  • Minutes

    Records and transcribes conversations locally; the agent searches their history.

  • Notion MCP

    Searches, reads and creates Notion pages and databases, and adds comments.

  • Octocode MCP

    Researches code on GitHub and npm – repositories, pull requests and packages – and optionally local code.

  • Olostep MCP

    Searches the web, scrapes and crawls pages and answers questions with sources.

  • OpenAI Docs MCP

    Searches and reads up-to-date OpenAI developer documentation: the API, Codex and models.

  • Oxylabs MCP

    Scrapes any page plus Google and Amazon results and drives a browser with AI.

  • PaddleOCR MCP

    Recognizes text and parses documents with tables, formulas and handwriting.

  • pdf-mcp

    Searches a PDF or a whole folder and cites source pages; reads tables and scans.

  • Qdrant MCP

    Stores information in Qdrant and finds it again with semantic search, as agent memory.

  • Rephonic MCP

    Data on millions of podcasts: audience size, demographics, contacts, charts and sponsors.

  • ScrapeGraph MCP

    Extracts data from pages based on a description, crawls sites and tracks changes on pages.

  • Scrapfly MCP

    Scrapes pages with proxies and JS rendering and drives a Scrapfly cloud browser.

  • Scrapling MCP

    Scrapes bot-protected pages through Scrapling’s stealth browser.

  • SE Ranking MCP

    Pulls keywords, backlinks, domain and SERP data, audits and rankings from SE Ranking.

  • Semrush MCP

    Analyzes keywords, domains and competitor traffic in Semrush and reads project data.

  • SEO Skill (seoskill.dev)

    Gives the agent your data from Search Console, Google Analytics, Bing and a site crawl.

  • Serpstat MCP

    Pulls domain data, keywords, competitors and backlinks from Serpstat for SEO analysis.

  • Similarweb MCP

    Checks website traffic and engagement and benchmarks sites against competitors.

  • Sosumi

    Searches and reads Apple documentation, Human Interface Guidelines and WWDC transcripts.

  • TwelveLabs MCP

    Indexes videos and finds scenes by description or image; summarizes and analyzes footage.

  • YouTube MCP Server

    Searches YouTube videos and channels, fetches transcripts and finds creators.

  • YouTube Transcript MCP

    Fetches transcripts of YouTube videos and Shorts, with timestamps and without sponsor segments.

  • yt-dlp MCP

    Downloads video, audio, subtitles and metadata from YouTube and other sites via yt-dlp.

Plugins (3)

  • Enterprise Search

    One question searches your company’s email, chat, documents and wiki at once.

  • Browser Use

    Gives the agent a real browser – your Chrome or a browser in Browser Use Cloud.

  • Claude Video Vision

    Lets Claude Code watch a video: frames via ffmpeg and a timestamped transcript.