compose-expert
A skill that gives Claude Code and other AI coding tools guidance for Jetpack Compose and Compose Multiplatform UI work across Android, Desktop, iOS, and Web, covering state, navigation, and animation.
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 that gives Claude Code and other AI coding tools guidance for Jetpack Compose and Compose Multiplatform UI work across Android, Desktop, iOS, and Web, covering state, navigation, and animation.
Claude Code skills that turn a video into a Chinese narration recap, covering scene detection, speech recognition, script writing, text-to-speech and ffmpeg assembly. Optional export to an editable JianYing or CapCut draft.
Agent skills that let Claude design mobile app screens with Sleek, edit Sleek projects, and implement those designs in React Native or HTML.
Claude Code skills that turn a video into a Chinese narration recap, covering scene detection, speech recognition, script writing, text-to-speech and ffmpeg assembly. Optional export to an editable JianYing or CapCut draft.
Push it to GitHub and send us the link. We read it, then it goes live.
A plugin of skills and agents for orchestrating Claude Code sessions. It covers watching pull requests through CI and review comments, filing reproducible bug reports, writing design briefs, and exploring codebases to locate files and trace behaviour.
A curated collection of 22 plugins for Claude Code, providing specialized agents, skills, and automation tools for development and productivity workflows.
A set of Claude Code skills that act as tutors for lifelong learning, covering school subjects, exam prep, languages and career skills, built on cognitive science methods such as active recall and Socratic dialogue.
Skills and agents for Claude Code covering Azure CLI management, Azure retail price lookups, Jira command-line workflows, and arbitrary-precision calculation. Useful for cloud, pricing, and ticket tasks.
Skills that generate privacy policies, terms of service, consent modals and cookie banners for websites and apps. Covers Korean, Japanese, EU and global rules, for Next.js projects.
A spec-driven development workflow for Claude Code that uses specialized sub-agents working in coordinated phases to turn a project idea into code. Useful for developers who want structured, multi-step AI-assisted builds.
Share, read, and continue agent sessions. Spool turns work done with coding agents into durable web pages that other people can understand and resume.
Reviews Swift code for concurrency mistakes, such as misused async/await, actors, Sendable and task groups, and points out modern API choices. Useful when you read, write or review concurrency code and want fewer bugs before they reach runtime.
Papr is a fast, native desktop RSS reader that works offline, with no account or cloud required. It also ships a papr command line tool that lets an AI agent read, search and triage your subscriptions.
A plugin of skills for auditing and improving AI agent workflows, covering prompt engineering, context management, tool orchestration, and guardrails. It helps teams speed up, adapt, and align their agent workflows.
High-performance job queue for Bun. SQLite by default, PostgreSQL multi-broker when you scale. DLQ, cron, SQLite S3 backups, and native MCP. No Redis.
Gives your coding agent working knowledge of current Solana development, covering dapp and wallet connection in Kit, Anchor and Pinocchio programs, tests with Surfpool and LiteSVM, client code generation, and common security pitfalls.
Gives a repository its own agent instructions and checks a change before handoff. Coordinates parallel investigation, edits and tests, and delegates small code lookups, edits and diff reviews to focused subagents, with compact commit messages.
Structural, AST-aware code navigation CLI for large, multi-language repositories.
A status line plugin for Claude Code that shows usage monitoring, context, rate limits, and cost tracking, across Claude, Codex, Gemini, Antigravity, and z.ai CLIs, with a modular widget system.
A memory layer that lets AI agents read and write an agent-owned second brain in an Obsidian vault. It works through MCP servers, lifecycle hooks and a command-line tool, and includes skills for recording preferences and setting up semantic search.
Skills for building iOS apps that use Meta AI glasses through the Wearables Device Access Toolkit. They cover camera streaming, photo capture, audio streaming, Swift conventions, and live SDK documentation search.
Transform YouTube videos from your favorite channels into beautifully formatted EPUB ebooks.
Gives indie developers one shared local desktop workspace where several coding agents can plan, hand off and build work together, with a visible plan and a record that ties agent output to project tasks. The agents you already run stay in your hands.
Gives you a personal AI agent that carries out multi-step work on your own computer, such as live research, organizing files and running scheduled tasks. Your memory, schedules and credentials stay local, and its tools can be shared with other agents.
Standalone agent skills for Claude Code that cover system design, architecture documentation, decision records and diagrams. Each skill keeps its scope narrow, such as recording one decision or documenting current architecture without proposing a target.
Red-teaming skills that set up an AI agent or chatbot target and run adversary attacks against it, then generate a report. Covers prompt injection and MCP server security checks.
Give Claude Code photographic memory in ONE portable file. No database, no SQLite, no ChromaDB - just a single .mv2 file you can git commit, scp, or share. Native Rust core with sub-ms operations.
Angular skills for Claude Code covering standalone components, dependency injection, directives, signal forms and HTTP data fetching. The skills are unmaintained, so use the official Angular skills instead.
Gives AI coding agents ready-made software development standards for accessibility, API design, architecture review and agent guardrails. Sync once and your agent follows them in any runtime, without you writing the rules into each prompt.
Watch Skill lets AI agents analyze video, audio and screen recordings, turning them into searchable, timestamped evidence. It answers questions about a recording and extracts chapters or bug reports from it.
A marketplace of hooks for Claude Code covering safety, automation and notifications, with skills to show repo TODO debt as bounties, session prompt-cache cost, token-heavy files, abandoned approaches and CLAUDE.md rule compliance.
Let AI agents debug your code inside VS Code - set breakpoints, step through execution, inspect variables, and evaluate expressions.
Walks a feature from a rough idea through requirements, design and implementation, with separate skills for debugging, codebase analysis and documentation. Helps a solo builder ship faster with less back-and-forth on what to build.
A Claude Code plugin that interviews you about what to build, designs the architecture, and writes a self-contained blueprint another Claude Code instance can build from with no prior context. Each build step includes acceptance criteria and a verify command.
Searches your own local PDF, PPTX and DOCX files offline. Use it to find information within your private documents, not for web research.
Helps you write novels, short stories and serial fiction. It drafts scenes in your own voice, tracks continuity across the story, and gives you structured critique and simulated reader reactions before you revise.
An autopilot skill that builds an app, site, bot, or feature from a spoken idea: it asks only where the idea forks, writes the spec, splits it into tasks, builds it and checks the result against the original brief, with live progress on a dashboard.
A methodology toolkit that turns a book on grassroots Chinese governance into a reusable skill, explaining county and township power structures and framing personal choices such as study, civil service exams, investment, and startups. It is a method guide, not the book's text.
Finds engaging moments in long videos, generates transcripts, and cuts highlight clips with titles and cover images. Built for creators who want short clips from long recordings.
A marketing skill covering website and landing-page audits with weighted scores, copywriting with panel scoring, an ad hook generator, and guidance on AI search visibility.
Reduces context and token usage when Claude Code reads, searches, or edits files, using cached file reads and a shared local knowledge graph. Useful when large files are re-read or the context window is filling up.
Workflow orchestration for multi-agent AI development, with an adaptive lifecycle engine, a knowledge graph, and skills for creating and tuning Claude Code skills. It is aimed at developers building multi-step AI workflows.
A Claude Code plugin for Elixir and Phoenix development, with skills for Ecto patterns, Oban jobs, LiveView, and dependency audits, plus contributor tools for tracking Claude Code changes.
Gives Claude ready-made workflows for everyday office work such as contracts, invoices, proposals, HR documents, reports, presentations and emails, so you skip writing the instructions yourself. Covers data analysis and document processing too.
A skill for professional readings of traditional Chinese fortune-telling, covering Zhouyi divination, astrology, date selection, home feng shui and fortune slips. It picks a method by question, subject and time range, then checks chart facts in layers to produce a self-contained reading brief.
A job hunt toolkit of separate skills for finding roles, decoding job descriptions, tailoring resumes, preparing behavioural interviews, comparing offers, and negotiating salary. Each skill works on its own.
Generates illustrations for articles, batch information slides for presentations, and cover images from a text input. It can output only the image prompt or a full image, and includes a Bento Grid style for feature displays.
Lets an AI agent or developer schedule and publish social media posts from the command line, across many platforms at once, without logging into each one. Also handles media uploads for those posts.
Gives Claude Code deep, domain-specific guidance for security work, such as reconnaissance, vulnerability scanning, reverse engineering and malware analysis, so you can get expert-level help on these tasks without writing the method yourself.
Lets an AI agent drive a local Firefox browser to read pages, fill forms and walk through multi-step website flows. It can also open the page you are building and inspect its console, network requests and performance so you can debug it without leaving your chat.
Mirrors the visual style of a reference paper figure onto your own data. Give it messy data and a figure screenshot, and it produces a matching plot for scientific papers.
Bridges WeChat messages into Claude Code through the iLink bot API. Lets you log in by scanning a QR code, check channel status, and manage who can message Claude and how direct messages are handled.
A Claude Code plugin of Harbor skills for building and evaluating LLM agents on long-horizon terminal tasks. It covers creating benchmark adapters, writing tasks and verifiers, and publishing tasks or datasets to the Harbor registry.
A thinking-partner skill that analyzes problems, decisions, and cost structures from the perspective of Elon Musk, drawing on his biographies, interviews, and public record. Invoked by phrases such as what would Musk think or Musk perspective.
Skills from ClickHouse that guide AI agents on ClickHouse schemas, queries and configuration, and run SQL or pandas-style analysis with chDB, an in-process ClickHouse engine for Python. Works with open-source ClickHouse and ClickHouse Cloud.
Turns a feature idea into a written spec, design and task list, then carries out the tasks one by one in a loop, so you can hand off a whole feature without losing track of the plan. Works with Claude Code and Codex.
A skills discovery plugin for Claude Code that searches for and installs Agent Skills, giving the agent specialized capabilities for tasks it would otherwise handle from base knowledge.
Launches, monitors and stops LLM inference workloads on one or more NVIDIA DGX Spark machines, so you can serve models across a cluster without Slurm or Kubernetes. Covers installing the tool and setting up the cluster, and managing the recipes that define each model run.
Decompiles Android APK, XAPK, JAR and AAR files with jadx or Vineflower, instruments apps with Frida, extracts HTTP API endpoints, and traces JNI and native library code and request signing. Aimed at security researchers doing authorized analysis.
Lets an AI assistant read and change your Todoist tasks on your behalf, so you can add, update and organize to-dos by describing them in conversation instead of clicking through the app. Runs as an MCP server or as tools inside your own projects.
An autonomous pipeline that turns a text description into a playable Godot game, generating design docs, game assets such as backgrounds, characters, props, cards and effects, and running visual checks and playtests.
Helps customize and extend Fireact SaaS apps after installation, such as adding features, pages and customizations. It detects Fireact projects automatically and works with Firebase authentication, Stripe billing, React and TypeScript.
A skill for Claude Code and Codex that runs SSH workflows through one Python CLI, covering remote commands, file transfer between machines, batch operations, tunnels and connection management on Windows, macOS and Linux.
A multi-agent code audit plugin that scans for security vulnerabilities, logic errors, and runtime bugs, then applies minimal fixes on a separate branch. It also reviews commit diffs and maps trust boundaries for threat modeling.
Gives you a guide to setting up Claude Code, covering global instructions, MCP servers, custom commands, hooks and skills, plus tools for generating commit messages and auditing code for security issues. This version is marked obsolete, with a newer starter kit recommended instead.
Runs a team of AI coding agents, such as Claude Code, Codex and Gemini CLI workers, in parallel from one dashboard or your phone. They share a task board, message each other, recover from crashes, and keep working when you are away.
Designs UI mockups for apps, admin panels, game interfaces and full landing pages, then turns them or screenshots into HTML/CSS, existing front-end projects or WeChat mini programs.
A set of Claude Code skills for operating Kurtosis, covering cleaning up enclaves, switching between Docker and Kubernetes backends, managing contexts, building the CLI locally, and debugging Docker engine and service issues.
Ratel provides skills for a Claude Code user to update per-package CHANGELOG.md files for a release and to publish local skills and loops to Ametyst, with optional retirement of the local originals.
Refines product specification documents by having multiple LLMs such as GPT, Gemini and Grok debate the draft until they reach consensus. Useful when a single model would miss gaps in a spec.
Runs recurring business work from inside Claude Code, with ready-made agents and skills for generating images, creating new agents and commands, and setting up goals and scheduled heartbeats through guided prompts, so you skip writing each markdown file by hand.
Gives you a practical handbook and ready-made skills for building, evaluating, and securing LLM agent systems. Its skills write architecture decision records, API design reviews, dataset profiles, postmortems, and pull request descriptions.
Persistent long-term memory for AI agents, shared across a fleet under access control. Agents recall prior decisions at the start of a task, with trust tiers, audit trails, and a knowledge graph.
A terminal pixel-art office where each AI coding agent session appears as a sprite, with skills for adding new agent CLI sources, redesigning office decorations, and running local code reviews.
Lets Claude Code run selected agents and models in parallel through the mco command line tool, compare their raw answers, and coordinate code review or implementation workflows.
Creates single-file HTML presentations that stay editable in the browser after generation. Supports object-level layout editing, slide reordering, local save and export, and converting PPT or PDF files into editable web slides.
A skill pack for security audits and code review, including a check of GitHub Actions that run AI agents for prompt injection and sandbox gaps. It also covers feature architecture design, project rules files, and frontend design guidance.
Connects Obsidian to Claude Code so meeting transcripts from Granola and notebooks from NotebookLM land as linked notes you can search. It also recalls past sessions from your vault memory, so you can pick up where you left off.
Gives your coding agent step-by-step guidance for common Dart tasks, such as writing unit tests and mocks, collecting coverage reports, building command-line apps, and fixing runtime errors from live stack traces.
Gives Claude Code and other AI coding agents a design system, with skills for applying brand context, fetching free placeholder assets, and critiquing UI quality. Free and open source under the MIT license.
Skills and agents for working on HyperShift, which hosts OpenShift control planes. Covers building and deploying control-plane and hypershift operator images, creating AWS hosted clusters for testing, and debugging hosted-cluster issues.
Verifies the authenticity of AI-agent authorization receipts and human device signoffs, including WebAuthn and passkey approvals. Useful when a user shares a trust receipt, authorization receipt, or signoff to check whether it is genuine.
Publishes content and articles to social media platforms by automating a browser through agent-browser, for users of AI coding assistants like Claude Code.
Skills and MCP tools that let an AI coding agent create avatar videos, giving it a digital face and voice so it can send video messages. Works with Claude Code and other AI coding agents.
A collection of agents for academic writing, covering peer review, literature review and paper revision. It grew out of the author's own paper-writing workflow.
A Claude Code plugin that gives a development team AI roles and workflow skills covering architecture design, data and API design, technology research, backend API development, and testing guides.
Gives Claude Code a team of specialized agents that plan, build, review, test and learn from past mistakes, so you can hand off a whole software task instead of directing each step yourself.
A Claude Code skill that grounds AI-generated UI in real reference images and a persistent per-developer taste profile, instead of generic AI-slop defaults.
A web research skill for Claude Code and other AI agents that searches the web through a self-hosted SearXNG instance and browses or scrapes pages with Camofox and CloakBrowser.
A command line tool for Reddit that lets Claude Code browse feeds, read posts, search, view users, upvote, save, and subscribe. It works through a reverse-engineered API.
Agent skills for building with ElevenLabs developer products, covering voice agent configuration, outbound calls, widget embedding and the procedure API. Works with any AI coding assistant that supports the Agent Skills specification.
Skills for searching Google Scholar with filters, tracking citing papers, finding full-text links, and exporting references to Zotero via BibTeX, driven through Chrome DevTools MCP.
Persistent, governed long-term memory for AI agent fleets, shared across agents under access control, with audit trails and a knowledge graph. Agents recall prior decisions at the start of tasks and reuse what earlier work established.
Lets Claude research how production Rails apps solve architectural problems, drawing on the Real World Rails repository of open source codebases. Useful for finding patterns or comparing approaches.
Lets React developers build SAP Fiori-style interfaces with ready-made components, instead of wiring up UI5 Web Components by hand. It also includes guidance for upgrading UI5 Web Components versions and for working with the analytical table internals.
Gives you an AI agent in a desktop app that works on your files, documents, browser, inbox and scheduled jobs. You approve each step before anything leaves your workspace, and your memory and history stay on your machine.
Rules\skills\subagens for vibecoding in 1C(bsl)
Lets Claude read, summarize, and log your Hevy workouts, routines, exercises, and body measurements from the terminal or through an MCP connection. Useful for reviewing training history or planning sessions.
Gives Claude Code a web search skill that queries a self-hosted SearxNG instance, aggregating Google, Bing, DuckDuckGo and Brave results. No API key is needed.
A Claude Code skill for designing mobile app screens, flows, and components, and for improving existing mobile interfaces, following established design principles.
A World of Tanks-style armored combat simulator built directly with Three.js and Vite: 220 vehicles, 33 destructible battlefields, and 13 modes, with plate-level armor, ballistics, modules, spotting, and physics. Play in your browser on desktop or mobile.
Official agentic skills for Qt software development and quality assurance, designed for use with AI coding tools such as Claude Code, Codex CLI, and GitHub Copilot. Skills have been tested with frontier LLMs from the Claude, Gemini, and GPT model families.
A ui-ux skill for Claude Code that builds and polishes app interfaces such as dashboards, lists, tables, forms, settings pages and modals. It works from your project's own component library and colors, and proposes wireframes for you to choose from before building.
Claude Code plugin for Butterbase, a backend-as-a-service for AI-native apps. It gives Claude knowledge of Butterbase's MCP tools and guides you through building apps, configuring auth, designing schemas, and deploying.
Turns a logo, some text or an emoji into the favicons, app icons and social sharing images a website needs, in every size required. Ask in plain language and the files are generated, checked and their HTML tags offered for your framework.
Gives Claude Code agents a structured specification language called Allium. Skills let you write, distil from existing code, elicit requirements into, and generate tests from Allium specs.
An in-the-wild benchmark for AI agents in the production harness.
Security skills for Claude Code from Trail of Bits, covering web fuzzing with ffuf, headless Ghidra binary reverse engineering, and threat modeling and vulnerability scanning, plus writing and research helpers.
Judges whether something actually produces value in a concrete scenario, tracking the evidence behind it and auditing the assumptions it depends on. Useful when you model a value scenario or question a figure an analysis relies on.
Taskmaster is a completion guard for coding agents. It addresses a common failure mode: the agent makes partial progress, writes a summary, and stops before the user goal is actually finished.
Turning data into a visual language for better thinking. AntV, initiated by Ant Group and open-sourced starting in 2017, reimagines data visualization by embedding the theory of graphical grammar into the JavaScript language.
Helps build production-grade Laravel REST APIs using the essa/api-tool-kit patterns, covering standard API responses, query filtering, dynamic pagination, and DDD-style endpoint scaffolding. Also covers general Laravel practices like dependency injection, auth, and testing.
A Claude Code skill for Shioaji, the SinoPac trading API for Taiwan markets. It helps with Python bindings, the shioaji CLI, HTTP API and streaming for order execution and market data.
Skills for Claude Code covering Linux administration, data platforms, feature stores and RAG pipelines, cloud networking, and security architecture. Each skill guides design and implementation work with AI assistance.
A Claude Code skill for creating, assigning, and monitoring software tasks through Machinist's GitHub issue workflow, intake labels, direct runs, and managed queue.
Lets your AI assistant speak and make media: chat in the voice of a real or fictional person, build music videos from songs and lyrics, generate sound effects, and turn the day's news into a spoken podcast briefing.
Fetches transcripts and metadata from YouTube, Instagram, TikTok, X and Vimeo, then saves an Obsidian-ready Markdown note with chapters and provenance. Can also summarize what was said in a video.
Generates cinematic AI short film prompts for tools such as Seedance, Sora, Kling, Veo and Jimeng, using a structure of stages. Useful for creators who want to turn a story idea into a detailed text prompt for video generation.
Gives an AI agent trading analyst skills for crypto and DeFi, covering market data, on-chain wallet research, backtesting and risk checks, so you can run multi-step market analysis in one conversation.
© 2026 Flowy
Built for Claude Code · Not affiliated with Anthropic
Logos are trademarks of their owners.