# Agent Skill Hub (full text)

> Open-source Agent Skills, Claude Code and Cursor plugins, MCP servers, agent harnesses and CLIs for AI coding agents.

## agent-browser

- Page: https://agent-skill-hub.pages.dev/e/agent-browser/
- Source: https://github.com/vercel-labs/agent-browser
- Kind: CLI for agents
- Author: Vercel Labs
- License: Apache-2.0
- Works with: claude-code, codex, cursor
- Tags: browser, cli, token-efficient
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `npm install -g agent-browser
agent-browser install`

Native Rust CLI that gives an agent a persistent browser: snapshot returns the page as compact text with element refs, and the agent acts on refs one command at a time.

Why it is useful: Shell-friendly browser automation with domain allowlists and an action policy, a good fit for lightweight scheduled agents.

## AGENTS.md

- Page: https://agent-skill-hub.pages.dev/e/agents-md/
- Source: https://github.com/openai/agents.md
- Kind: Spec or format
- Author: OpenAI and contributors
- License: MIT
- Works with: codex, cursor, claude-code, opencode
- Tags: agents-md, format, context
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

A simple open format, a README for agents, that gives coding agents a predictable place for project context and instructions.

Why it is useful: One instruction file that most harnesses read, so project rules do not have to be copied into each tool's config.

## Anthropic Agent Skills

- Page: https://agent-skill-hub.pages.dev/e/anthropic-skills/
- Source: https://github.com/anthropics/skills
- Kind: Skill pack
- Author: Anthropic
- License: Mixed: Apache-2.0 for example skills, proprietary for document skills
- Works with: claude-code, claude-ai, claude-api
- Tags: agent-skills, official, documents, skill-authoring
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `/plugin marketplace add anthropics/skills`

Anthropic's public Agent Skills: document skills (docx, pdf, pptx, xlsx), skill-creator, mcp-builder, frontend-design, webapp-testing and more, plus the Agent Skills spec and a template.

Why it is useful: The reference implementation of the SKILL.md format. Read skill-creator and the template before writing your own skill, and install the document skills when an agent must produce Office files.

## Awesome Claude Code

- Page: https://agent-skill-hub.pages.dev/e/awesome-claude-code/
- Source: https://github.com/hesreallyhim/awesome-claude-code
- Kind: Curated list
- Author: hesreallyhim
- License: CC BY-NC-ND 4.0
- Works with: claude-code
- Tags: awesome-list, discovery
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

A hand-picked list of Claude Code resources: skills, hooks, slash commands, status lines, orchestrators and applications, with a recently-added section.

Why it is useful: Good discovery feed for new Claude Code tooling; check each linked project's own license before reuse.

## Awesome Claude Code Subagents

- Page: https://agent-skill-hub.pages.dev/e/awesome-claude-code-subagents/
- Source: https://github.com/VoltAgent/awesome-claude-code-subagents
- Kind: Curated list
- Author: VoltAgent
- License: MIT
- Works with: claude-code
- Tags: subagents, awesome-list
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

A collection of Claude Code subagent definitions for specific development tasks, installable as a Claude Code plugin.

Why it is useful: Ready-made subagent prompts to copy or adapt when a task needs a narrow specialist with its own context.

## Awesome Claude Skills (Composio)

- Page: https://agent-skill-hub.pages.dev/e/awesome-claude-skills-composio/
- Source: https://github.com/ComposioHQ/awesome-claude-skills
- Kind: Curated list
- Author: Composio
- License: No top-level license file
- Works with: claude-code, codex, cursor
- Tags: awesome-list, agent-skills, discovery
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

A large list of Claude Skills and plugins, with some skills stored in the repo (skill-creator, mcp-builder, file-organizer, content-research-writer and others).

Why it is useful: Broad coverage of non-coding skills; parts of the README promote Composio's own gateway, so judge each skill on its own.

## Awesome Claude Skills (travisvn)

- Page: https://agent-skill-hub.pages.dev/e/awesome-claude-skills-travisvn/
- Source: https://github.com/travisvn/awesome-claude-skills
- Kind: Curated list
- Author: travisvn
- License: No license file
- Works with: claude-code, claude-ai
- Tags: awesome-list, agent-skills, discovery
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

A curated list of Claude Skills, resources and tools, with an explanation of how skills load through progressive disclosure.

Why it is useful: A concise primer plus links for people new to Agent Skills.

## Awesome MCP Servers

- Page: https://agent-skill-hub.pages.dev/e/awesome-mcp-servers/
- Source: https://github.com/punkpeye/awesome-mcp-servers
- Kind: Curated list
- Author: punkpeye
- License: MIT
- Works with: any-mcp-client
- Tags: awesome-list, mcp, discovery
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

A curated list of Model Context Protocol servers grouped by category, plus clients and other MCP resources.

Why it is useful: The quickest way to check whether an MCP server already exists for a service before writing one.

## Browser Use

- Page: https://agent-skill-hub.pages.dev/e/browser-use/
- Source: https://github.com/browser-use/browser-use
- Kind: Library
- Author: Browser Use
- License: MIT
- Works with: python
- Tags: browser, agent-loop, python
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Open-source browser agent library that runs locally with any LLM; a paid cloud adds hosted browsers.

Why it is useful: Use it when a task needs a full agent loop over web pages rather than single commands.

## ccusage

- Page: https://agent-skill-hub.pages.dev/e/ccusage/
- Source: https://github.com/ryoppippi/ccusage
- Kind: CLI for agents
- Author: ryoppippi
- License: MIT
- Works with: claude-code, codex
- Tags: cost, observability, cli
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `npx ccusage@latest`

CLI that analyses coding-agent token usage and cost from local session data, with daily, monthly and session reports.

Why it is useful: Measure what a daily agent actually costs before and after you change its prompt or schedule.

## Chrome DevTools MCP

- Page: https://agent-skill-hub.pages.dev/e/chrome-devtools-mcp/
- Source: https://github.com/ChromeDevTools/chrome-devtools-mcp
- Kind: MCP server
- Author: Chrome DevTools team
- License: Apache-2.0
- Works with: claude-code, cursor, codex, any-mcp-client
- Tags: browser, mcp, debugging, performance
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Lets a coding agent control and inspect a live Chrome through DevTools, as an MCP server or a CLI, for automation, debugging and performance traces.

Why it is useful: Gives agents real DevTools data (network, console, performance) when checking front-end work; category flags limit the tool set.

## Claude Agent SDK for Python

- Page: https://agent-skill-hub.pages.dev/e/claude-agent-sdk-python/
- Source: https://github.com/anthropics/claude-agent-sdk-python
- Kind: Agent SDK
- Author: Anthropic
- License: MIT
- Works with: claude-code
- Tags: sdk, python, official, agent-loop
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `pip install claude-agent-sdk`

Python SDK that exposes Claude Code's agent loop, tools, hooks and MCP support so you can build your own agents in Python 3.10+.

Why it is useful: Use it when a scheduled agent needs custom tools or structured control in Python instead of shelling out to the CLI.

## Claude Code

- Page: https://agent-skill-hub.pages.dev/e/claude-code/
- Source: https://github.com/anthropics/claude-code
- Kind: Agent harness
- Author: Anthropic
- License: Proprietary (Anthropic Commercial Terms); repo is public
- Works with: claude-code
- Tags: coding-agent, terminal, official, headless
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `npm install -g @anthropic-ai/claude-code`

Anthropic's agentic coding tool for the terminal, IDE and GitHub. The public repository holds the changelog, issue tracker and example plugins.

Why it is useful: The harness most Agent Skills and plugins in this list target. Its headless print mode makes it a scheduler-friendly runner for small daily agents.

## Claude Code Plugins Directory

- Page: https://agent-skill-hub.pages.dev/e/claude-plugins-official/
- Source: https://github.com/anthropics/claude-plugins-official
- Kind: Plugin marketplace
- Author: Anthropic
- License: Apache-2.0
- Works with: claude-code
- Tags: plugins, official, marketplace, code-review
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Anthropic's curated Claude Code plugin directory: internal plugins such as code-review, feature-dev, hookify, plugin-dev, LSP integrations and ralph-loop, plus partner plugins under external_plugins.

Why it is useful: A vetted starting point for plugins and a working example of how plugins bundle skills, commands, agents, hooks and MCP servers.

## Claude Code Router

- Page: https://agent-skill-hub.pages.dev/e/claude-code-router/
- Source: https://github.com/musistudio/claude-code-router
- Kind: Model gateway
- Author: musistudio
- License: MIT
- Works with: claude-code, codex, opencode, pi
- Tags: gateway, routing, multi-provider
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Local model gateway and control plane that gives Claude Code, Codex, OpenCode, Pi and other clients one endpoint, with routing, failover and observability across providers.

Why it is useful: Run the same agent workflows on cheaper or local models and fail over when one provider is rate limited.

## Claude Code Templates

- Page: https://agent-skill-hub.pages.dev/e/claude-code-templates/
- Source: https://github.com/davila7/claude-code-templates
- Kind: Curated list
- Author: Daniel Avila
- License: MIT
- Works with: claude-code
- Tags: templates, hooks, discovery
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

An open catalog (aitmpl.com) of about 1,900 Claude Code components: skills, agents, slash commands, MCP integrations, settings, hooks and project templates, installable with one command.

Why it is useful: Fast way to browse and install components; review what each installs, since the catalog mixes official and community content.

## Cloudflare MCP Servers

- Page: https://agent-skill-hub.pages.dev/e/cloudflare-mcp-server/
- Source: https://github.com/cloudflare/mcp-server-cloudflare
- Kind: MCP server
- Author: Cloudflare
- License: Apache-2.0
- Works with: claude-code, cursor, any-mcp-client
- Tags: cloud, deploy, mcp, official
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Cloudflare's domain-specific remote MCP servers (docs, Workers bindings, observability and more) over Streamable HTTP; the README recommends the newer Cloudflare API MCP server.

Why it is useful: Lets an agent read and change Cloudflare configuration, useful when deploying static sites to Cloudflare Pages.

## Codex CLI

- Page: https://agent-skill-hub.pages.dev/e/codex-cli/
- Source: https://github.com/openai/codex
- Kind: Agent harness
- Author: OpenAI
- License: Apache-2.0
- Works with: codex
- Tags: coding-agent, terminal, official
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

OpenAI's open-source coding agent for the terminal, with IDE integrations; supports AGENTS.md and skills.

Why it is useful: A second harness for cross-checking another agent's work or running reviews with a different model family.

## Context7

- Page: https://agent-skill-hub.pages.dev/e/context7/
- Source: https://github.com/upstash/context7
- Kind: MCP server
- Author: Upstash
- License: MIT
- Works with: claude-code, cursor, codex, any-mcp-client
- Tags: docs, mcp, context
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

MCP server and CLI that fetch current, version-specific library documentation and code examples into the agent's context.

Why it is useful: Cuts hallucinated or outdated API usage when an agent writes code against fast-moving libraries.

## Continual Learning (Cursor plugin)

- Page: https://agent-skill-hub.pages.dev/e/cursor-continual-learning/
- Source: https://github.com/cursor/plugins/tree/main/continual-learning
- Kind: Plugin
- Author: Eric Zakariasson
- License: MIT
- Works with: cursor
- Tags: memory, agents-md, hooks
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Cursor plugin that keeps AGENTS.md up to date from transcript changes, using a stop hook plus a continual-learning skill that writes short high-signal bullets.

Why it is useful: Turns repeated corrections into durable project memory without hand-editing AGENTS.md after every session.

## Cursor Plugins

- Page: https://agent-skill-hub.pages.dev/e/cursor-plugins/
- Source: https://github.com/cursor/plugins
- Kind: Plugin marketplace
- Author: Cursor
- License: Per plugin (first-party plugins checked: MIT)
- Works with: cursor, claude-code
- Tags: plugins, official, marketplace
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Cursor's official multi-plugin marketplace: pstack, thermos, ralph-loop, continual-learning, orchestrate, advisor, cursor-team-kit and integrations such as GitHub, Playwright and X.

Why it is useful: Each plugin is a standalone folder with a manifest, so it doubles as a reference for building Cursor plugins; the repo also works as a Claude Code marketplace.

## ego lite (ego-browser skill)

- Page: https://agent-skill-hub.pages.dev/e/ego-lite/
- Source: https://github.com/citrolabs/ego-lite
- Kind: Skill
- Author: CitroLabs
- License: MIT
- Works with: claude-code, codex, cursor
- Tags: browser, agent-skills, macos
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `npx skills add citrolabs/ego-lite`

The ego-browser agent skill and its open-source Node.js CDP harness for the ego lite browser: the agent runs ego-browser skill for a version-matched guide, then sends JavaScript that calls snapshot, fill, click, wait, navigate and capture inside its own isolated Space.

Why it is useful: For agents that need the user's real browser logins without taking over the user's tabs, with parallel Spaces per agent. Caveat: the ego lite browser app itself is closed source; only the harness and skill are in this repo, and it runs on macOS and Windows only.

## Firecrawl MCP Server

- Page: https://agent-skill-hub.pages.dev/e/firecrawl-mcp-server/
- Source: https://github.com/firecrawl/firecrawl-mcp-server
- Kind: MCP server
- Author: Firecrawl
- License: MIT
- Works with: claude-code, cursor, any-mcp-client
- Tags: web, scraping, mcp, research
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

MCP server for Firecrawl: search, scrape and interact with the live web and get clean Markdown or structured data back.

Why it is useful: Turns arbitrary pages into LLM-ready text for research agents; uses the Firecrawl API, so it needs a key.

## GitHub MCP Server

- Page: https://agent-skill-hub.pages.dev/e/github-mcp-server/
- Source: https://github.com/github/github-mcp-server
- Kind: MCP server
- Author: GitHub
- License: MIT
- Works with: claude-code, cursor, codex, any-mcp-client
- Tags: github, mcp, official
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

GitHub's official MCP server for reading repositories and code, managing issues and pull requests, and monitoring Actions runs.

Why it is useful: Gives an agent scoped GitHub access through toolsets instead of ad-hoc API calls.

## gstack

- Page: https://agent-skill-hub.pages.dev/e/gstack/
- Source: https://github.com/garrytan/gstack
- Kind: Skill pack
- Author: Garry Tan
- License: MIT
- Works with: claude-code, codex, cursor
- Tags: workflow, code-review, qa, browser, planning
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Garry Tan's skill pack that turns Claude Code and other agents into a virtual engineering team: office hours, CEO and eng plan reviews, design review, code review, QA with a real browser, security audit and ship.

Why it is useful: Opinionated, battle-tested slash-command workflows for the whole loop from idea to deploy; the /browse and /qa skills give an agent a headless browser for checking its own work.

## HyperFrames

- Page: https://agent-skill-hub.pages.dev/e/hyperframes/
- Source: https://github.com/heygen-com/hyperframes
- Kind: Library
- Author: HeyGen
- License: Apache-2.0
- Works with: claude-code, codex, cursor
- Tags: video, rendering, agent-skills
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `npx skills add heygen-com/hyperframes`

Open-source framework that renders HTML, CSS, media and seekable animations into deterministic MP4 video, with agent skills for Claude Code, Codex and Cursor.

Why it is useful: Lets a coding agent make explainer videos by writing HTML scenes, while rendering stays deterministic.

## lark-cli

- Page: https://agent-skill-hub.pages.dev/e/lark-cli/
- Source: https://github.com/larksuite/cli
- Kind: CLI for agents
- Author: Lark / Feishu
- License: MIT
- Works with: claude-code, codex, cursor
- Tags: feishu, productivity, agent-skills, official
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Official Lark and Feishu CLI built for humans and agents, covering Messenger, Docs, Base, Sheets, Calendar, Mail, Tasks and more, with 24 bundled agent skills.

Why it is useful: Lets an agent read and write Feishu docs and Bitable from the shell, for example docs +fetch to pull a research note into a pipeline.

## LinkedIn MCP Server

- Page: https://agent-skill-hub.pages.dev/e/linkedin-mcp-server/
- Source: https://github.com/stickerdaniel/linkedin-mcp-server
- Kind: MCP server
- Author: Daniel Sticker
- License: Apache-2.0
- Works with: claude-code, cursor, codex, any-mcp-client
- Tags: mcp, browser, python
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `uvx mcp-server-linkedin@latest --login`

Python MCP server that drives LinkedIn through the user's own browser session with Patchright: tools read profiles, companies, employees, jobs, the feed, posts and the inbox, and a smaller set sends messages, connection requests and job dismissals.

Why it is useful: For job-search and networking agents that need LinkedIn reads from the user's own session. Caveat from the README: LinkedIn's User Agreement prohibits automated access and accounts can be restricted, so deny the write tools (send_message, connect_with_person, reply_to_conversation, archive_conversation, dismiss_job) and keep volume low.

## live-panel

- Page: https://agent-skill-hub.pages.dev/e/live-panel/
- Source: https://github.com/ythx-101/live-panel-skill
- Kind: Skill
- Author: ythx-101
- License: MIT
- Works with: claude-code, codex
- Tags: video, rendering, agent-skills
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `git clone https://github.com/ythx-101/live-panel-skill ~/.claude/skills/live-panel`

Agent skill and stdlib Python renderer that turns one JSON config into an always-running terminal-style or pastel architecture diagram, with flowing packets, a scrolling log and counters, rendered to an H.264 mp4 or a self-contained looping web page.

Why it is useful: For agents that make animated architecture or agent-tree diagrams for posts and videos; the agent writes the JSON, renders, then runs check_frames.py to catch text overflow. Needs Chrome or Chromium and ffmpeg; Linux and macOS only (Windows through WSL). Example designs credit their original authors.

## magpie

- Page: https://agent-skill-hub.pages.dev/e/magpie/
- Source: https://github.com/yetone/magpie
- Kind: Model gateway
- Author: yetone
- License: MIT
- Works with: claude-code, codex, gemini-cli, opencode
- Tags: gateway, routing, macos, multi-provider
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Menu-bar app and local gateway that serves every coding agent's model from one place and moves to the next account when a quota runs out.

Why it is useful: Keeps scheduled agents running across several subscriptions and providers without editing each tool's config.

## MCP Inspector

- Page: https://agent-skill-hub.pages.dev/e/mcp-inspector/
- Source: https://github.com/modelcontextprotocol/inspector
- Kind: CLI for agents
- Author: Model Context Protocol
- License: MIT
- Works with: any-mcp-client
- Tags: mcp, testing, official
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `npx @modelcontextprotocol/inspector`

Developer tool for inspecting MCP servers through a web UI, a scriptable CLI, a terminal UI and an experimental connection daemon.

Why it is useful: Its CLI lets you test an MCP server in CI or from an agent before wiring it into a harness.

## MCP reference servers

- Page: https://agent-skill-hub.pages.dev/e/mcp-reference-servers/
- Source: https://github.com/modelcontextprotocol/servers
- Kind: MCP server
- Author: Model Context Protocol
- License: MIT
- Works with: any-mcp-client
- Tags: mcp, reference, official
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Reference MCP server implementations maintained by the MCP steering group: everything, fetch, filesystem, git, memory, sequential thinking and time.

Why it is useful: Small, readable examples of correct MCP servers; the README points to the MCP Registry for production servers.

## MCP Registry

- Page: https://agent-skill-hub.pages.dev/e/mcp-registry/
- Source: https://github.com/modelcontextprotocol/registry
- Kind: Curated list
- Author: Model Context Protocol
- License: MIT
- Works with: any-mcp-client
- Tags: mcp, registry, official, api
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

The official registry service that gives MCP clients a list of MCP servers, with a public API at registry.modelcontextprotocol.io.

Why it is useful: A machine-readable source for discovering and publishing MCP servers, suitable for automated catalog updates.

## OpenCode

- Page: https://agent-skill-hub.pages.dev/e/opencode/
- Source: https://github.com/anomalyco/opencode
- Kind: Agent harness
- Author: Anomaly
- License: MIT
- Works with: opencode
- Tags: coding-agent, terminal, multi-provider
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `npm i -g @opencode/cli@latest`

Open-source terminal coding agent that works with many model providers, with a desktop app and an npm CLI.

Why it is useful: Provider-neutral harness for running skills and plans on models other than the default vendor's.

## Pi

- Page: https://agent-skill-hub.pages.dev/e/pi-mono/
- Source: https://github.com/badlogic/pi-mono
- Kind: Agent harness
- Author: Mario Zechner (badlogic)
- License: MIT
- Works with: pi
- Tags: coding-agent, terminal, extensible, headless
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

A minimal, extensible coding-agent harness: customise it with extensions, skills, prompt templates and themes; run it interactively, in print or JSON mode, over RPC or through its TypeScript SDK.

Why it is useful: A small harness you can read and bend, good for scheduled agents that need predictable non-interactive runs.

## Playwright CLI

- Page: https://agent-skill-hub.pages.dev/e/playwright-cli/
- Source: https://github.com/microsoft/playwright-cli
- Kind: CLI for agents
- Author: Microsoft
- License: Apache-2.0
- Works with: claude-code, codex, cursor
- Tags: browser, cli, token-efficient
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Playwright command-line interface shipped with skills, aimed at coding agents that prefer concise CLI calls over large MCP tool schemas.

Why it is useful: Lower token cost than the MCP server for agents that only need a few browser actions per task.

## Playwright MCP

- Page: https://agent-skill-hub.pages.dev/e/playwright-mcp/
- Source: https://github.com/microsoft/playwright-mcp
- Kind: MCP server
- Author: Microsoft
- License: Apache-2.0
- Works with: claude-code, cursor, codex, any-mcp-client
- Tags: browser, mcp, testing
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

MCP server that drives a browser with Playwright through structured accessibility snapshots, so models act on pages without screenshots.

Why it is useful: A standard browser tool for any MCP client; the README itself says it is not a security boundary, so enforce allowed origins outside it.

## pstack

- Page: https://agent-skill-hub.pages.dev/e/pstack/
- Source: https://github.com/cursor/plugins/tree/main/pstack
- Kind: Plugin
- Author: Lauren Tan (poteto)
- License: MIT
- Works with: cursor, claude-code
- Tags: workflow, verification, code-quality, principles
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote
- Install: `/plugin marketplace add cursor/plugins
/plugin install pstack@cursor-plugins`

poteto's Cursor plugin of rigorous engineering skills (poteto-mode, interrogate, tdd, create-verification-skill, no-comments and about 30 principle skills), also installable in Claude Code.

Why it is useful: Pushes agents toward less but verified code: verification skills, root-cause fixes and principle checks make parallel agents trustworthy.

## Ralph Loop (Cursor plugin)

- Page: https://agent-skill-hub.pages.dev/e/cursor-ralph-loop/
- Source: https://github.com/cursor/plugins/tree/main/ralph-loop
- Kind: Plugin
- Author: Cursor
- License: MIT
- Works with: cursor
- Tags: loop, hooks, autonomy
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Runs Cursor in a self-referential loop with afterAgentResponse and stop hooks, feeding the same prompt back until a completion promise appears or an iteration cap is hit.

Why it is useful: A compact, readable implementation of the Ralph Wiggum technique for long tasks that need many unattended iterations.

## Serena

- Page: https://agent-skill-hub.pages.dev/e/serena/
- Source: https://github.com/oraios/serena
- Kind: MCP server
- Author: Oraios AI
- License: MIT
- Works with: claude-code, cursor, codex, any-mcp-client
- Tags: code-navigation, mcp, lsp
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

MCP server that gives agents IDE-like semantic tools: symbol-level code retrieval, editing, refactoring and debugging through language servers.

Why it is useful: Symbol-aware edits on large codebases use fewer tokens and break less than line-based edits.

## Spec Kit

- Page: https://agent-skill-hub.pages.dev/e/spec-kit/
- Source: https://github.com/github/spec-kit
- Kind: CLI for agents
- Author: GitHub
- License: MIT
- Works with: claude-code, codex, cursor, copilot
- Tags: workflow, planning, spec-driven
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

Toolkit that gives coding agents structured processes and templates, starting with spec-driven development, plus extensions for bug fixing and assessment.

Why it is useful: A repeatable spec, plan and task flow that keeps agent work reviewable on larger features.

## Stagehand

- Page: https://agent-skill-hub.pages.dev/e/stagehand/
- Source: https://github.com/browserbase/stagehand
- Kind: Library
- Author: Browserbase
- License: MIT
- Works with: typescript
- Tags: browser, extraction, typescript
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

TypeScript browser automation framework that mixes code with natural-language act, extract and observe steps, and runs a local browser or Browserbase.

Why it is useful: Schema-validated extract() replaces brittle selectors for structured scraping from pages that change layout.

## Superpowers

- Page: https://agent-skill-hub.pages.dev/e/superpowers/
- Source: https://github.com/obra/superpowers
- Kind: Skill pack
- Author: Jesse Vincent (obra)
- License: MIT
- Works with: claude-code, codex, cursor, gemini-cli, opencode
- Tags: workflow, tdd, debugging, planning
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

A software development methodology delivered as composable skills: brainstorming, writing and executing plans, test-driven development, systematic debugging, subagent-driven development and verification before completion.

Why it is useful: Gives an agent a disciplined default process; the skills are short and easy to read, fork or cherry-pick.

## wshobson/agents

- Page: https://agent-skill-hub.pages.dev/e/wshobson-agents/
- Source: https://github.com/wshobson/agents
- Kind: Plugin marketplace
- Author: Seth Hobson
- License: MIT
- Works with: claude-code, codex, cursor, opencode, pi
- Tags: subagents, plugins, multi-harness
- Added: 2026-11-04; last verified 2026-11-04 via git ls-remote

One catalog of 94 plugins with about 200 agents, 185 skills and 105 commands, adapted for Claude Code, Codex CLI, Cursor, OpenCode, GitHub Copilot and Pi.

Why it is useful: A large, multi-harness source of specialist agents and skills with a capability matrix that shows what works where.