paperorchestra
A skill pack that runs the PaperOrchestra multi-agent pipeline inside coding agents such as Claude Code and Cursor, turning raw research notes into a LaTeX paper ready for submission.
4,248 plugins found in public marketplaces and GitHub repositories, indexed and searchable. Plugin skills also install in Codex, Cursor and 4 more agents. In Claude Code, the 99 that get Auto-invokedWhat is this?This plugin ships a FLOW.md router the engine fires, so the matching skill runs itself. No slash command to remember.Learn how → fire as you prompt.
A skill pack that runs the PaperOrchestra multi-agent pipeline inside coding agents such as Claude Code and Cursor, turning raw research notes into a LaTeX paper ready for submission.
Gives you a self-hosted runtime for running autonomous AI agents and swarms that delegate work, keep durable memory, and run on schedules. It connects to Claude, GPT, Gemini, Ollama and other model providers.
Evidence-based agent skills that guide a design-award application end to end: finding comparable award winners, evaluating a design against a transparent rubric, matching it to eligible award programs, preparing entry text, and checking the submission.
Turns a rough topic into a confirmed brief, then writes platform-specific copy and short-video scripts for YouTube, Douyin, Xiaohongshu and WeChat. It also analyses post performance and saves each finished draft, with you approving direction, titles and final versions first.
Push it to GitHub and send us the link. We read it, then it goes live.
Gives your AI agent ready-made instructions for marketing and growth work, such as finding cold-email addresses, writing blog cover images, naming a brand, and mining app reviews for copy ideas. Install the skills you need so you stop doing these tasks by hand.
A local knowledge base that Claude Code and other agents share, keeping decisions, meeting notes, SOPs, debugging lessons and preferences across sessions. Search and save notes, organize files and run workflows from the terminal.
A skill for Claude Code and Codex that helps prevent Terraform and OpenTofu mistakes by catching identity churn, secret exposure, blast-radius errors, CI drift, and gaps in compliance gates.
A lightweight framework for building AI agents, with a data analysis skill that loads, transforms, aggregates, charts and exports tabular data such as CSV, JSON and Parquet using Polars.
Helps Claude design and polish user interfaces by applying the concrete rules from the book Refactoring UI, such as set spacing, type, color and shadow scales and building visual hierarchy with weight and color. It turns vague design judgment into specific choices you can apply directly.
A motion design studio on your own machine: Claude makes the video step by step, you approve every stage. Bring your own Claude Code or API key.
Skills for running a supervised network of Claude Code or Codex agents that collaborate on your machine. Covers the wanman CLI for inter-agent messaging and task management, plus artifact naming, quality and cross-validation standards for research and structured deliverables.
A WeChat bridge that lets you chat with Claude Code from WeChat. Scan a QR code to link your account, then messages you send are forwarded to Claude Code on your computer and replies stream back, with support for text, images, voice and files.
A Claude Code plugin marketplace with skills for bulk code refactoring, transferring code between files, file metadata analysis, and running Python locally for multi-file operations. It also offers parallel ensemble solving for architecture decisions.
Adjusts how third-party models behave inside Claude Code so they run closer to a native Opus session, covering subagent lifecycle, session hooks and tool interception. It does not replace your gateway, provider mapping or account setup.
Breaks a build goal into phased plans and recon steps for autonomous coding. Fires when you ask Claude to plan or run a multi-step task across a codebase.
A skill for data scientists and analysts that writes cost-safe geospatial SQL for BigQuery, DuckDB, Snowflake, Wherobots, and Postgres, and can render results on an interactive map. Works locally or self-hosted without a SaaS account.
A command-line interface for Firecrawl that lets you search, scrape, crawl, map and download websites, and run autonomous data extraction jobs from your terminal. It also searches developer and documentation sources.
Gives Claude Code a skill for biomedical literature and variant research using the BioMCP command-line tool. It also lets you file issues about the tool itself in the biomcp repo.
Skills that let Claude control WPS Office through natural language, covering Excel formulas, data cleaning and charts, Word formatting and editing, PowerPoint layout and animations, and cross-app tasks.
Lets your AI coding assistant jump to definitions, find every usage of a symbol, and rename across a codebase, using the language servers you already run. It works even when the assistant gets line and column numbers wrong.
You wanted to learn negotiation. You got a 40-hour course, a syllabus someone wrote years ago, and a tab you never reopened. Claudemy is a free Claude Code skill that builds the course you will actually finish. You name a topic.
One prompt in, a looping animated 3D icon out (with real transparency) an animated WebP you can drop straight into your app UI.
An AI-powered Claude skill for creating, rigging, and refining Spine 2D skeletal animations, from raw assets to interactive previews. Live Animation Preview · Interactive Part Editor.
Builds draw.io architecture diagrams with AWS, GCP, and Azure stencils and a layout engine. Fires when you ask for a cloud architecture diagram or a Kubernetes topology in draw.io.
Skills for Claude Code that help scientists draft, verify and revise research manuscripts, covering citation checks, claim-source verification, data availability statements, and removing defensive or self-critical wording. Some skills also support figure planning and reviewer responses.
13 Claude Code skills for launch-grade motion graphics. Every frame is code. No After Effects, no video generator.
Gives AI agents access to React Native DevTools data and Rozenite plugins on a live app, through shell commands or the agent SDK. Covers network, storage and performance inspection.
Autonomous GitHub Project build → QA → review loop for Claude Code. One-line install.
Audits any website for SEO, technical, Core Web Vitals, structured data, security, and AI search readiness using the SEOmator CLI. Returns prioritized reports with health scores across its rule set.
Helps AI produce distinctive interface designs instead of generic templates. It previews visual directions, implements the chosen one, and checks it in the browser, and it also makes cover images for social media platforms.
Flutter and Dart skills that a coding agent loads on its own, covering architecture, Bloc state management, Dart idioms, accessibility, and code review. Each skill is a folder with a SKILL.md file.
Gives a founder or marketer ready-made help with copy for ads, landing pages and emails, competitor research, conversion advice for landing pages, and go-to-market plans, so they stop starting each prompt from scratch.
Gives Claude Code design guidance for pages, components and product interfaces, covering diagnosis, visual direction previews, design systems, implementation and visual checks. It also creates and adjusts cover images for WeChat, Bilibili, YouTube and Xiaohongshu.
Easy Authentication and Authorization with Keycloak in .NET.
Gives a marketing team AI-agent skills for planning a period's marketing, building content calendars, writing campaign briefs, diagnosing weak campaign numbers, and drafting short-form video scripts. Covers US, EU, SEA and LATAM markets in Vietnamese and English.
A slide-making skill that builds presentation-grade .pptx decks from your actual work, using only facts from your source material, with fully editable PowerPoint output and a review pass before delivery.
A Claude Code plugin for planning and producing AI-generated music albums, with skills for album concepts, tracklists, progress dashboards, idea tracking, and album artwork prompts.
A house style for hand-written SVG diagrams such as flowcharts and architecture figures, with layout arithmetic, a colour system, and a dependency-free linter that checks the output. Suited to CJK-friendly typography.
Exposes local services through secure tunnels, giving public HTTPS, TCP, or UDP URLs for testing webhooks, demos, and deployments. Built in Rust and billed pay-as-you-go.
An agent harness and toolkit for building AI agents and agentic applications, with a CLI and SDKs. It supports sub-agent spawning, MCP integration, session memory and human-in-the-loop control.
Gives your AI coding agent expert Rust guidelines for ownership, error handling, async and concurrency, so it writes idiomatic code instead of clone-heavy, unwrap-everywhere code. Works with most agents that support skills.
A secure\* runtime for autonomous AI agents, where security policy is derived from a human-readable constitution. \*When someone writes "secure," you should immediately be skeptical. What do we mean by secure?
Vestige is a fail-closed runtime firewall for AI agents. The model proposes, a deterministic gate decides, and every action leaves a receipt. When something breaks, causal_walk finds the real cause.
A Claude Code skill that finds compelling moments in a video, such as funny dialogue or repeated impact actions, and cuts them into standalone clips. It can also reformat clips between widescreen and vertical formats for social media.
Create, edit, and implement Figma designs with your coding agent.
Open Reality: phone video to AI-queryable 3D scenes. MCP tools for Claude/Codex/Cursor (npm openreality-mcp), the self-hostable broker, and the VGGT-SLAM library, in one repo. BSD-2-Clause.
Gives Claude Code live access to web search, URL scraping, news, social media, market prices and on-chain crypto data. Useful when a task needs current information beyond the model's training data.
Lets robotics teams build simulated benchmark scenes and test robot policies in them at scale, without hand-writing a separate setup for each robot, object and scenario combination. Covers environment authoring, test runs and pulling evaluation results.
An open source X/Twitter automation toolkit with an MCP server for AI agents, a CLI and browser scripts. It covers account backup, scraping, analytics, unfollow management and engagement automation.
A CLI tool that lets coding agents message, watch, and spawn each other across terminals. Start an agent with hcom in front, then prompt normally.
Gives your AI coding assistant ready-made guidance for browser automation, AI SDK questions, real-time messaging servers, current library docs, and improving a skill's prompt. Each one walks the agent through the task without you re-explaining it.
Helps AI job seekers prepare for interviews with a Chinese-language cheat sheet on ML and LLM topics, plus a skill that generates an academic personal homepage from a CV and a skill that renders research artifacts as readable HTML.
Helps Claude Code write, convert, and review YAMLScript (.ys) files, including converting Clojure into idiomatic YAMLScript following established style rules.
Copies AI chat conversations from ChatGPT, Claude, Gemini and similar tools as clean Markdown context bundles, with one-click copying from the sidebar and all processing done locally.
Bridge between Claude and OpenAI's Codex CLI, get AI-powered code analysis, generation, and review right in your editor.
An agent skill that gives a coding agent color science knowledge for palettes, ramps, gradients, color space conversions, accessibility checks, and perceptual color matching.
Connects Claude Code to WhatsApp and Instagram Business messaging through the OpenBSP platform, with tools for managing allowed contacts, webhooks and account setup.
A skill for Claude Code that gives the agent persistent memory of its mistakes. The agent maintains a markdown file in your repo (.claude/napkin.md) where it tracks what went wrong, what you corrected, and what worked.
An open source AI video workbench that Claude Code can drive over MCP, with storyboarding, reference images, keyframe generation and an editable timeline. It runs locally with any model or your own ComfyUI setup.
Overture is an open-source MCP server with a local web interface that draws an AI coding agent's execution plan as an interactive flowchart before the agent writes code. It lets you review the plan visually first.
Lynkr is an LLM gateway that routes requests across many providers, with code intelligence and compression features to reduce token use in Claude Code.
Gives AI coding agents access to Shopify's developer documentation, API schemas, and code validation for building apps and themes, plus store management through the Shopify CLI. Also includes a skill for finding and buying products from UCP merchants.
Skills that give Claude Code guidance for building interfaces that follow Apple's Human Interface Guidelines for iPhone, iPad, Mac and Apple TV, plus Material Design for Android. They cover components, navigation, input and accessibility.
Turns a complete beginner's vague app idea into a buildable plan, then keeps them oriented while they build with AI coding tools. Aimed at people who have never coded and do not yet know what to build.
Lets MCP-compatible AI agents deploy apps to Google Cloud Run, with a skill for managing Cloud Run services and jobs. Works from AI coding CLIs, IDEs, and agent SDKs.
Avibe is a local-first agent workbench that lets you drive Claude Code, Codex and OpenCode from your browser or chat apps. Its skills cover durable delegation, background watches, secure credential use and prompt auditing.
Extracts news and article content from sites such as WeChat public accounts, Toutiao, Netease, Tencent, BBC, CNN, Twitter/X and Quora, returning clean JSON or Markdown for Claude Code users.
Interact with a Marmot data catalog from Claude Code to search data assets, view lineage, browse glossary terms, check pipeline runs, and manage tags and owners. Gives AI agents access to catalog metadata for tables, topics, queues and APIs.
Open-source AI film studio · agent + node workflows Direct AI films, shot by shot. Start with one sentence. The agent writes the story, draws a still for every shot, and cuts the clips on a timeline.
Lets you create, connect, list and inspect secure tunnels that expose a private or localhost MCP server through a local tunnel client. Useful for reaching MCP servers that are not publicly reachable.
Helps investors judge whether a stock's price fairly reflects its real growth, and drafts source-backed buy-side research memos from a ticker. Also covers valuation trend health and market-cycle stage analysis.
Guides a feature from idea to working code in Claude Code, with agents drafting the artifacts for each stage while you review and approve before the next one begins.
A Claude Code skill for Roblox developers using the Nevermore Luau engine. It converts a Luau file from nonstrict to strict type checking, adding type annotations and fixing the type errors the checker reports.
A Claude Code skill that guides the agent to write code in the style of a senior engineer, applying SOLID principles, test-driven development, and clean code practices when coding, refactoring, designing, or reviewing.
A user-invoked Claude Code skill that audits an entire codebase for tech debt and architecture problems. It writes TECH_DEBT_AUDIT.md with file-cited findings, severity ratings, and effort estimates.
Adds skills for creating Cursor rules, Agent Skills, and subagents, plus converting older rules and commands to the Agent Skills format and editing Cursor or VSCode settings. Works within the Cursor editor.
Lets AI agents operate TouchDesigner projects directly, so they can create, change and delete nodes, read project structure and run Python inside a running TouchDesigner session. Includes skills for debugging node errors and for checking integration tests before a release.
Adds feature flags, A/B tests, and gradual rollouts to Next.js, SvelteKit, or other apps. Includes skills for declaring flags with the Flags SDK and wiring up a provider adapter.
Source code and evaluation framework for the agent skills that are included with Claude for Teachers.
A plugin of skills and agents for Claude Code covering the development workflow, from exploring the codebase and writing briefs to filing bugs and babysitting pull requests through CI and review.
Applies the classic writing rules from Strunk's style guide to any prose people will read, such as documentation, commit messages, error messages and reports, so the text comes out clearer and more concise.
Compiles a repository's instruction files such as AGENTS.md and CLAUDE.md into a rubric that checks code changes against your project rules, then validates and calibrates it. Useful for enforcing coding standards in Claude Code.
A meta-skill that builds project skills from a codebase's rules, workflows, and lessons, gathering them into a skills directory that AI coding agents such as Claude Code and Cursor can consult before each task.
Helps cloud teams adopt FinOps by giving them customizable tools for analyzing Microsoft Cloud cost data with KQL queries and FinOps hubs. It also offers guidance on Microsoft writing style for customer-facing content.
A cyber threat intelligence and OSINT analysis toolkit that runs structured investigations and produces analyst-style reports with sourced, trust-scored findings. Built for security analysts working OSINT and CTI cases in Claude Code.
A persistent, file-based memory layer that gives Claude Code and other coding agents a project brain for durable decisions, requirements and constraints, managed through a zero-dependency CLI.
A local CLI and MCP server that gives AI agents SEO tools, reading your own Search Console, Google Analytics, Bing Webmaster Tools and Clicky data, plus keyword research from DataForSEO and Ahrefs.
Operate a Coolify self-hosted PaaS from Claude through an MCP server, letting you inspect, deploy, restart, and debug apps in plain English. Destructive actions wait for human confirmation.
A modern desktop application for managing Claude Code configuration files.
A Claude Code plugin for video editing that analyzes footage, transcribes audio, builds contact sheets, and generates XML edit timelines. It also includes skills for backing up and restoring libraries.
Bootstraps a no-code Postgres backend for an app you describe, with schema and project setup. Fires when you start a new app or ask to scaffold a database for generated code.
Traditional Chinese skills for AI agents with local file and terminal access. They set up a video editing environment and turn a user's own footage into vertical short videos, protecting original media and checking outputs before reporting them done.
Chat-driven biomedical research assistant for Claude Code that runs bioinformatics, RNA-seq, drug discovery and clinical analysis workflows, with a live dashboard and Feishu rich-card reporting for progress updates.
A plugin for creating, improving, and testing Claude Code skills. It includes a skill-creator skill that runs evals to measure how well a skill performs.
Skills for AI education and content creation: adaptive tutoring and study guides, multi-lesson course artifacts, single-file HTML AI/ML survey papers, image generation and editing, and an LLM council that has several open-weight models rank each other's answers before a final synthesis.
Checks whether your brand or domain gets cited in AI answers from ChatGPT and AI search, tracks how often you are mentioned against competitors, and suggests which fixes to make first. Uses live public data, read-only, so it researches but never posts.
Provides skills that run deterministic compiled graphs inside the ZeroClaw, Hermes Agent and OpenClaw agent runtimes, with importers and MCP tools. It also tracks AI co-authored code attribution in projects.
A Claude Code plugin of DuckDB skills for querying, exploring, and converting data files, plus session memory for attached databases. Covers SQL queries, file format conversion, extension installs, and DuckDB documentation search.
Control Figma from the command line to create shapes, frames, text and components, set styles and layout, manage variables, and export images. Useful for automating design tasks in Claude Code.
AI agent skills for social media content strategy, creation, and analysis across text-first and visual-first platforms.
Audits an agent skill before you trust it, flagging shell commands, network calls, tool invocations and credential risks it could carry. Run it inside Claude Code with the semia scan command on a skill path.
Manages personal local Agent Skill files, detecting reusable workflows during or after sessions so they can be saved and reused later. Also includes a universal release workflow that detects version files and changelogs across Node.js, Python, Rust and Claude plugin projects.
A Claude Code plugin that provides colgrep, a local semantic code search tool that can replace Grep and Glob for finding code by meaning. It uses multi-vector ColBERT embeddings with optional regex filtering.
Claude Code skills for offensive security work, covering penetration testing, bug bounty hunting, AI threat testing, API security, authentication attacks and multi-hop attack path mapping. Built for security testers working within authorized scope.
Extracts article content from any URL as clean Markdown, including JavaScript-rendered pages and WeChat articles, with headings and links preserved. Useful for turning web pages into text for LLM workflows.
An open-source Claude Code plugin that turns a one-line goal into a product requirements document through a PM-style interview, then breaks it into a dependency-ordered task graph and executes each task with verification.
Generates illustrations for documents, articles and notes by analysing their structure, extracting the key points and creating a matching image for each topic. Supports cover images and custom aspect ratios.
A Chinese-language translation of Anthropic's Skills repository, covering skills for algorithmic art, brand guidelines, canvas design, co-authoring documents and working with Word (.docx) files. It helps Chinese-speaking developers learn Claude skills and use the document-processing guides.
A writing skill that steers Claude toward natural, human-sounding prose by avoiding common AI writing patterns. It applies to tweets, emails, articles, bios, captions, reports, and cover letters.
A cross-platform computer repair skill for Windows, macOS and Linux, usable from AI coding agents. It helps diagnose and clean up a full system drive, slow performance, malware, app migration and data recovery, gathering evidence first and changing things only after confirmation.
Runs a PDCA workflow with phase gating, spec checks, and delegation to sub-agents for planning, QA, and deployment. Fires at session start and on prompts, edits, task creation, and phase transitions.
Gives separate Claude Code, Codex and similar agent sessions one shared local workspace, where they coordinate in channels and threads, track work on task boards, and keep memory so you stop re-explaining context. Can also turn a task into a reviewable HTML page.
Orchestrates Claude Code, Codex, Antigravity, Grok Build and OpenCode behind one runtime, with persistent sessions, multi-agent councils and autonomous loops. Includes a skill that interviews the user to produce a full app spec for building a new app.
A Claude Code plugin that checks AI-generated code against its design specs, with audit logs, auth, and database skills for bkend.ai projects. Useful for developers reviewing code that an AI assistant wrote.
Docker-authored skills that help AI coding agents with Docker work, such as writing Dockerfiles, optimizing images, composing services, and running Docker Agent.
Skills for working with Elastic: onboard an Elastic Cloud organization and manage Serverless projects, traffic filters and user roles, create and manage Elasticsearch anomaly detection jobs, and diagnose non-green cluster health.
A Claude Code plugin for designing modular software architectures and reviewing existing code for coupling problems, using the Balanced Coupling model. It produces design and review documents in Markdown and HTML.
© 2026 Flowy
Built for Claude Code · Not affiliated with Anthropic
Logos are trademarks of their owners.