omniroute
A free, MIT-licensed AI gateway that routes requests through one endpoint to many model providers, with quota-aware fallback and prompt compression. It ships CLI skills for chat, batch jobs, backups and agent tasks.
Flowy lists 176 plugins for Cloudflare work in Claude Code. Each one bundles skills a Claude Code agent can load while you work. Every listing shows what is inside before you install, who built it, and whether it is Auto-invoked, meaning a FLOW.md router fires the right skill as you prompt instead of leaving them dormant.
A free, MIT-licensed AI gateway that routes requests through one endpoint to many model providers, with quota-aware fallback and prompt compression. It ships CLI skills for chat, batch jobs, backups and agent tasks.
Skills for building, converting and migrating MCP Apps with interactive UI views, plus a skill that rebuilds a branch into a clean git commit history.
Gives Claude Code a Payload skill for working with Payload CMS projects, covering collections, fields, hooks, access control and the Payload API, plus commands for dependency audits, CI flake triage, test conversion, UI review, issue triage and translation generation.
Builds and tracks job-search materials for Claude Code users: a complete resume as importable Reactive Resume JSON, mock and prep interviews, a job application pipeline, and offer negotiation help.
Phaser is a fast, free, and fun open source HTML5 game framework that offers WebGL and Canvas rendering across desktop and mobile web browsers and has been actively developed for over 13 years.
Draw and edit draw.io diagrams by chatting with an AI. Describe a diagram in a sentence and the AI draws it on a real draw.io canvas. Edit it by hand like any draw.io file, or select a few shapes and tell the AI what to change.
Gives an AI agent the step-by-step workflow a senior security analyst follows, covering when to use each technique, what to check first, how to run it and how to verify the result. Covers incident response, forensics, threat intelligence, malware analysis, penetration testing and red teaming.
Open source SEO toolkit for Claude Code with skills for keyword clustering, competitor analysis, and checking how a brand appears in AI answers. Useful for marketers who find Semrush or Ahrefs too costly.
Lets you control Claude Code sessions remotely from a mobile app, with end-to-end encryption, push notifications and sync across devices. Open source under the MIT license.
n8n-MCP is an open-source MCP server that lets Claude, ChatGPT, Cursor, and other AI assistants operate n8n: read failed executions, edit workflows in place, validate and test them, manage credentials, and build new workflows.
A personal AI harness for Claude Code that keeps memory of your goals and routes requests to skills for research, scraping, writing, and visual content, helping you move from your current state toward your ideal one.
A personal AI infrastructure for Claude Code with reusable skills, agents and commands for scraping social and business data, searching arXiv papers, creating visuals, curating quotes, and running multi-pass analysis, aimed at individuals managing their life and work.
A full-stack TypeScript CMS built on Astro for serverless sites, with skills for building sites, creating plugins, command-line management, browser UI testing and adversarial code review.
Gives local dev servers stable named .localhost URLs in place of port numbers, and includes skills for setting up OAuth redirect URIs against those URLs. Useful for developers and agents working on local web apps.
A unified gateway for agent tools, giving an agent one base URL and one token to reach a catalog of endpoints across many providers. Covers SEO and backlinks, social and trends, and people and company enrichment.
A Model Context Protocol server that lets AI assistants control Home Assistant smart home devices in natural language, query device states, run services, and manage automations.
A pixel-art office dashboard that shows in real time what each AI assistant is working on, whether it is online, and what it did recently. It supports multiple agents, a Chinese, English and Japanese interface, and can be deployed as a standalone status board.
Gives a developer one shared set of agent rules and reusable workflow skills, covering tasks such as controlling your signed-in Chrome, redacting and previewing agent transcripts, and checking ClickClack deploys. Use it to keep several local coding agents working the same way.
A local-first Agent Workbench where people and agent teams share one workspace to run multiple agent engines, manage plugins and Skills, and coordinate multi-agent projects and tasks.
Analyzes codebases to find duplicate business logic, check coupling between modules, flag misplaced classes, size components, and plan migrations away from monoliths.
An MCP server that lets Claude create and open diagrams, flowcharts, and mockups in the draw.io editor.
A skill that scaffolds, extends, and verifies reactive local-first JavaScript or TypeScript apps built on TinyBase, a small in-memory store for tabular and key-value state with sync support.
Rapid-MLX is an open-source LLM inference server for Apple Silicon, built on MLX. It offers OpenAI- and Anthropic-compatible APIs with reliable tool calling for coding agents.
Chrome extension & CLI to let agents control your browser. Runs Playwright snippets in a stateful sandbox. Available as CLI or MCP.
Keeps long multi-step coding work on track inside Codex. It holds project context across sessions, plans the work, carries it out one step at a time, and checks the result before calling the task done.
Gives Claude Code a bash-style interface to files and directories mounted from cloud storage, databases, and APIs in one virtual filesystem. Includes skills for working with mounted data and for building custom adapters for new data sources.
PinMe is a zero-config deployment CLI focused on one-command creation and deployment for full-stack projects. It lets you quickly set up and launch a complete project with an integrated frontend, Worker backend, and database, without tedious configuration.
Lets an AI assistant read and change your Gmail, Drive, Calendar, Docs, Sheets and other Google Workspace apps in plain language, so you stop clicking through each app by hand. Works from Claude Code via MCP tools or a command line interface.
A local-first issue board you can run in a browser, with a bundled skill that lets Claude manage its issues and setup through the taskctl CLI.
A collection of Agent Skills for building on Cloudflare, Workers, the Agents SDK, and the wider Cloudflare Developer Platform.
Connects an AI agent to everyday chat apps such as WeChat, Feishu, Slack and Discord, so you can talk to it from your phone and keep it running around the clock. Runs locally and works with multiple models.
Loopy helps Claude Code users discover, compare, audit, run, and save repeatable AI-agent loops for recurring coding work. A companion skill to the Loop Library website.
Scaffolds documentation sites with Nuxt and Markdown, with support for multilingual docs, llms.txt output and an AI docs assistant. Includes a sample skill for testing agent skills discovery.
Gives you ready-made prompt recipes for hand-drawn and illustrated image styles, such as crayon, ink wash, pixel art and sketch looks, so you can get a consistent style without writing the prompt yourself.
Skills for building open data portals with PortalJS. Scaffold a portal, add datasets, charts and geospatial files, migrate an ArcGIS Hub site, and publish DCAT catalog feeds for national open-data harvesters.
Helps you plan and direct AI-generated films, from story structure and shot composition to camera action choreography and cinematic lens looks. It is built to keep long videos coherent across scenes, so you can stitch short generator clips into a longer movie.
Shows you what is slow, failing or expensive in your app, across traces, logs and metrics in one place. It is built on OpenTelemetry, so any OTLP exporter can send data to it, and it can also run as a local binary.
An adapter that lets Pi use MCP servers, with a scripting skill that covers tool discovery, how call results are shaped, and script limits. It is meant for developers who want MCP tools without filling the context window.
A plugin for automating routine work with AI models, with approval gates, scheduled routines, and a builder for small local microapps to manage the results. Open source and local-first.
Takes an app built by a coding agent from repository to live production on your own accounts. You choose the providers for hosting, database, auth, payments, email and domain setup, and connect the accounts yourself.
Runs your repository's tests, builds and UI checks on a remote machine while you keep editing locally. Choose a cloud box, a local VM or container, or a machine you already own.
A Claude Code plugin marketplace offering agent-based tools for development workflows, code analysis, code comparison, and code grading, plus skills for research, content, and residential floor plan searches.
Deploys and operates a self-hosted OpenTag setup for Slack, pairing a local runner with a GitHub-connected Control Plane so an engineering channel gets an AI teammate that stays reachable even when its computer is offline.
Gives a development team ready-made commands and agents for code review, testing, releases and deployment, plus skills for transcribing audio, extracting video frames and managing Cloudflare or BigCommerce work.
An MCP server that lets AI assistants connect to and manage Jupyter notebooks in real time, with kernel status and execution. It can run notebook code locally or on cloud sandboxes such as Datalayer, Kaggle, Google Colab and Modal.
Fetches, filters and analyzes real-time news and papers from many tech, finance and AI sources such as Hacker News, GitHub, arXiv and Hugging Face Papers. Built for Claude Code users who want a current digest of technology and AI news.
Agent skill for creative 3D visualization: turn an idea into a Three.js/Blender scene worth exploring. Claude Code + Codex plugin, 223 recipes, 440 knowledge records, 22 proved kits, 37 runnable studies.
A Claude Code plugin for AI-human collaboration across the development lifecycle, with skills for OpenSpec change implementation, archiving and exploration, frontend browser verification, and release blog writing.
Gives Claude Code, Codex and Cursor a tested setup of skills, commands, subagents and MCP servers, so you start with working guidelines that curb wrong assumptions and bloated code instead of configuring each tool yourself.
Get YouTube transcripts, search videos, browse channels, and extract playlists, from any AI agent. YouTube Skills gives your AI agent instant access to YouTube transcripts, video search, channel data, and playlist extraction.
Bernstein is a Claude Code plugin for orchestrating AI agents. It adds skills to list and inspect agents, review and approve tasks, create tasks, track costs and budgets, and show alerts for failed or stalled work.
A research toolkit for Claude Code that carries a scientific question through to a compiled LaTeX manuscript, with skills for publication-quality figures, narrative structure for reports, cross-project research memory, and friction sheets for computational physics and quantum tools.
Production workflow skills for Claude Code. Each skill guides Claude through a recipe to produce tangible output, scaffolded projects, generated assets, professional documents, deployed services. Ten plugins of practical, production-oriented skills.
A bug bounty agent framework for Claude Code that automates reconnaissance, vulnerability hunting and exploit chaining, with skills for analyzing recon output, running autonomous hunt loops and correlating findings into attack chains.
A local-first AI research workbench for scientific work. It searches the literature, runs Python and R, queries scientific databases, and keeps figures, runs, decisions, and drafts together in one project.
Manage UniFi Network, Protect and Access from Claude Code, with skills for setting up the UniFi Access server, investigating device incidents over a set time window, reviewing area activity, and preparing sanitized support bundles for bug reports.
An HTTP and REST based Redis client for TypeScript, built on the Upstash REST API. It suits serverless functions such as AWS Lambda, where persistent connections are impractical.
Lets you check Sentry errors from your terminal or editor, viewing issues, events and projects and making API calls without opening the web dashboard. Useful for triaging crashes and reading stack traces while you code.
An MCP server and plugin that lets an AI agent drive ComfyUI end to end: it checks locally installed models, builds the workflow graph, frees VRAM, and runs renders. Includes skills for training LoRAs and color correction.
Gives Claude Code a skill for red-teaming a web application under authorization. It drives pages, captures the HTTP traffic behind them, inventories what the target exposes, and replays mutated requests with auditable evidence.
Self-hostable YesCaptcha-compatible captcha solver built with FastAPI, Playwright, and OpenAI-compatible multimodal models.
Assigns issues from your Linear, GitHub, GitLab or Slack to an agent that works each one in its own Git worktree and reports progress back as it goes. Bring your own model keys or subscriptions.
An Android app that runs as an MCP server, letting AI models remotely control an Android device through accessibility services and screenshot capture.
Skill for building with yahoo-finance2, an unofficial TypeScript, Deno and npm library for Yahoo Finance market data. Covers its data modules, CLI and MCP server.
Gives AI agents persistent memory across sessions, with semantic search, a knowledge graph, and document storage backed by SQLite. Load stored context at session start, recall past work, and save important information for later.
Comprehensive Cloudflare platform reference docs for AI/LLM consumption. Covers Workers, Pages, storage (KV, D1, R2), AI (Workers AI, Vectorize, Agents SDK), networking, security, and infrastructure-as-code.
Skills for querying on-chain token, wallet and market data, scoring token creators and call-out wallets, and executing trades or token launches across Solana, BSC and Base.
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.
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.
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.
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.
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.
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.
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.
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.
Share, read, and continue agent sessions. Spool turns work done with coding agents into durable web pages that other people can understand and resume.
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 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.
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.
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.
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.
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.
Reads web pages, documents, images, audio and YouTube videos and converts them into clean, token-lean Markdown that you or your AI agents can use. Self-hosted.
Gives Claude Code agents a task board for planning, assigning, claiming, and reviewing multi-step project work. Includes skills for maintaining scheduled work, running whole project plans, delegating single tasks, and letting workers update their assigned tasks.
Claude Code skill for web scraping with scrapling - auto Fetcher selection, Cloudflare bypass, site patterns.
Reverse-engineers the design taste of a website from its URL. A real browser captures the page's structure and a screenshot, then analysis produces a write-up of its design tokens and the reasons behind its UI choices.
Helps you write AI video prompts that work, picks the right model for your clip, and translates or debugs prompts across Kling, Seedance, HappyHorse and other video models, so you stop trial-and-error regenerating.
Generates e-commerce product images and campaign photo sets with GPT-Image models through a user-configured OpenAI-compatible endpoint, using scene templates. It can also export prompt packs for manual use.
Lets an AI assistant work with your Odoo database through guarded tools, so agencies can review many client databases, check data quality before trusting answers, plan version upgrades, and run month-end accounting closes with human sign-off at each step.
Helps an architecture studio set up its own AI practice, covering how AI is governed in the studio, which practice areas it covers, and what it remembers. It gives architects ready-made skills for tasks like color palettes, FF&E schedule exports and site demographics.
Connect Claude to Korea's MOLIT real estate API and simulate buy now / buy later / invest only scenarios based on your income, savings, and retirement goals.
Go from game idea to deployed, monetized browser game in minutes, not hours. Tell your AI coding agent "make a space invaders game" and get a fully playable 2D or 3D browser game with pixel art, procedural audio, and automated QA. Deploy instantly to the web.
A command line tool and skills that let Claude Code drive a stealth browser for websites with bot detection, CAPTCHAs, or anti-bot blocks, covering page navigation and form filling.
Skills for building and running AI agents on OpenMA, a self-hostable platform for durable sessions, sandboxed tools, and encrypted credentials. Covers creating agents through conversation, deploying OpenMA, and managing its environments and vaults.
Gives Claude Code access to a Buddhist digital text platform for searching scripture and commentary across Chinese, Pali and Tibetan canons. Its sutra-commentary skill compares how classical commentators explained a passage.
Lets an AI assistant check Daiso and convenience store stock, compare product prices, and find nearby shops, Olive Young stores, and cinema showtimes and seats, so you can ask in plain language instead of searching each site by hand.
Runs equity research, earnings analysis, and sentiment checks on stocks. Fires when you ask to analyze, compare, or gauge market sentiment on a ticker.
Sync Claude Code sessions across devices using Cloudflare R2 with end-to-end encryption.
Gives Claude Code persistent memory across conversations, with local embeddings, semantic search and a knowledge graph. It runs offline as a single binary, with no cloud service or API keys needed.
A lightweight desktop tool that runs a local gateway for the Codex app, translating its Responses API requests into Chat Completions format and forwarding them to a provider you choose. A desktop UI manages providers, model mapping, ports and logs.
Exposes a local web app, static site, or service through a Portal relay so it can be reached publicly without port forwarding, firewall rules, or manual DNS setup. Also helps set up your own relay on a Linux host.
Gives your AI coding agent ready-made skills for dashboards and analytics, AWS accounts and agents, Python agent building, and Bun projects, so it handles these tasks with less back and forth. Works with Claude Code, Copilot, Cursor and Windsurf.
Bosun is an automation plugin for autonomous engineering work, covering workflow orchestration, pull request lifecycle handling, multi-agent routing and CI quality gates. It includes skills for Playwright browser testing and for controlling native Windows desktop apps.
Gives AI coding assistants ready-made instructions for scraping data from Apify Actors, covering ads, app stores, job boards and AI search visibility. Each skill tells the assistant which Actor to run, how to build its input, and how to recover when a scrape gets blocked.
Run multiple coding-agent sessions in parallel, and see which one needs you. A browser terminal for parallel AI coding agents: several sessions side by side, each in its own cell, with the one that needs you marked in colour.
This is the new v3 MCP server. The previous v2+ MCP server is deprecated and lives at dataforseo/mcp-server-typescript-deprecated. MCP server and CLI for LLM agents to browse DataForSEO API documentation and make authenticated API requests.
System audio capture + multi-provider ASR + local-first AI review workspace. Floating live captions, 12 ASR backends, 60+ languages, AI summary/chat/mindmap, Open API, MCP server, and Agent Skill.
A TypeScript framework for building MCP servers that enforce security in the architecture, with a typed egress firewall so raw data does not reach the LLM unfiltered. Includes skills for development and deploying to Vinkius Edge.
Gives each AI agent its own real email address and phone number, so it can send and receive email and text messages and place outbound voice calls without manual setup. Each agent gets an isolated inbox, and outbound messages are scanned so keys or personal data do not leak.
Open-source collection of skills compatible with the SKILL.md standard. Developed for the AI courses at UCEMA.
Clone, commit, diff, branch, merge, rebase, cherry-pick, stash, tag, and manage worktrees in local git repos via MCP. STDIO or Streamable HTTP.
Gives your AI agent persistent memory across sessions, so it stops forgetting past work and decisions. Knowledge is stored as linked memories recalled by association, with no separate database or embedding bill.
AI agent skills are reusable instruction sets that extend your assistant with domain-specific expertise, loaded on demand so they don't bloat your context. This repository covers marketing and engineering skills authored by @samber.
Gives developers ready-made guidance for building web apps and APIs, covering animated React components, authentication with JWT and OAuth, API design and contract testing, and changelog writing. Works inside Claude Code.