> ## Documentation Index > Fetch the complete documentation index at: https://usefulai.com/llms.txt > Use this file to discover all available pages before exploring further. # Best Plugins, Skills, and MCP Servers for AI Agents in 2026 > Find useful plugins, skills, and MCP servers for the way you work. Everything you can add to your AI agent, in one place - ranked by what people actually install. Scan the list, or head into a category when you know what you're looking for.
| # | Name | Category | Best for | Est. installs About install estimates | | --: | :----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :---------------------------------------------------------------------------------------------------------- | :------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------: | | 1 | GitHub | Coding | Issues, PRs, and CI on GitHub | **1.6M** | | 2 | Find Skills | Setup & Memory | Discovering skills you can install | **1.4M** | | 3 | Context7 | Coding | Current, version-correct library docs | **1.2M** | | 4 | Frontend Design | Design & UI | Escaping the generic AI look | **1.2M** | | 5 | Superpowers | Coding | Full engineering discipline end to end | **980K** | | 6 | Playwright | Coding | Browser flows and end-to-end tests | **790K** | | 7 | Azure | Infrastructure | Inspecting and operating Azure at scale | **740K** | | 8 | Grill Me | Setup & Memory | Grilling a plan until decisions are explicit | **670K** | | 9 | agent-browser | Automation | Local CLI browser built for agents | **600K** | | 10 | ECC | Coding | Adopting a whole team method at once | **600K** | | 11 | Chrome DevTools | Coding | Debugging and profiling live pages | **530K** | | 12 | Gmail | Communication | Search, summarize, and draft in Gmail | **510K** | | 13 | Google Workspace CLI | Productivity | Scriptable access to all of Workspace | **500K** | | 14 | Web Design Guidelines | Design & UI | Auditing UI against Vercel's rules | **490K** | | 15 | Code Review | Coding | Automated second pass on pull requests | **430K** | | 16 | Grill with Docs | Coding | Nailing requirements before building | **430K** | | 17 | Caveman | Setup & Memory | Terse replies, fewer output tokens | **410K** | | 18 | Skill Creator | Setup & Memory | Building and measuring your own skills | **390K** | | 19 | Notion | Productivity | Notion pages, databases, and capture | **350K** | | 20 | Supabase | Coding | Schema, SQL, and migrations on Supabase | **340K** | | 21 | Code Simplifier | Coding | Cleaning up freshly written code | **340K** | | 22 | Slack | Communication | Workspace context and channel digests | **320K** | | 23 | Google Calendar | Productivity | Scheduling against your real calendar | **320K** | | 24 | Ralph | Coding | Looping through small verifiable tasks | **310K** | | 25 | UI UX Pro Max | Design & UI | Concrete design-system recommendations | **300K** | | 26 | Vercel | Coding | Next.js and Vercel-stack development | **300K** | | 27 | Remotion | Media Creation | Programmatic video in React | **300K** | | 28 | Outlook Email | Communication | Inbox search and thread recaps | **300K** | | 29 | Microsoft Teams | Communication | Catching up on chats and channels | **300K** | | 30 | Anthropic PPTX | Design & UI | Producing and editing PowerPoint files | **300K** | | 31 | CLAUDE.md Management | Setup & Memory | CLAUDE.md files that stay current | **290K** | | 32 | TDD by Matt Pocock | Coding | Behavior-first test discipline | **290K** | | 33 | Taste Skill | Design & UI | A collection of distinct design workflows | **285K** | | 34 | Improve Codebase Architecture | Coding | Choosing the next high-value refactor | **280K** | | 35 | Feature Dev | Coding | Structuring one substantial feature | **280K** | | 36 | shadcn/ui | Design & UI | UI aligned with your shadcn components | **260K** | | 37 | n8n | Automation | Building and running n8n workflows | **250K** | | 38 | Cloudflare | Infrastructure | Workers and the Cloudflare platform | **250K** | | 39 | HyperFrames | Media Creation | Plan-render-review video pipeline | **240K** | | 40 | TypeScript LSP | Coding | Live TypeScript type errors | **240K** | | 41 | Security Guidance | Coding | Security review while you code | **240K** | | 42 | Writing Great Skills | Setup & Memory | A method for authoring skills | **240K** | | 43 | Claude Code Setup | Setup & Memory | A shortlist of automations for your repo | **230K** | | 44 | claude-mem | Setup & Memory | Session memory that persists locally | **220K** | | 45 | DeepWiki | Coding | Orienting in unfamiliar public repos | **220K** | | 46 | Impeccable | Design & UI | A disciplined design-critique loop | **210K** | | 47 | Firecrawl | Search & Web | Websites turned into agent-ready data | **210K** | | 48 | Storybook | Design & UI | UI built against your real components | **200K** | | 49 | Figma | Design & UI | Design-to-code from real Figma context | **200K** | | 50 | Docker MCP Toolkit | Infrastructure | Running many MCP servers in containers | **200K** | | 51 | Tavily | Search & Web | Search, extract, crawl, and research | **200K** | | 52 | Outlook Calendar | Productivity | Meeting prep and Outlook scheduling | **200K** | | 53 | Browser Use | Automation | General web browsing and form-filling | **190K** | | 54 | Postman | Coding | API testing from Postman workspaces | **190K** | | 55 | CodeRabbit | Coding | Independent review of AI-written code | **190K** | | 56 | Web Artifacts Builder | Design & UI | Polished self-contained web artifacts | **180K** | | 57 | Google Drive | Productivity | Working across your Drive documents | **170K** | | 58 | Desktop Commander | Automation | Broad control of your own computer | **170K** | | 59 | GitLab | Coding | Merge requests and CI on GitLab | **170K** | | 60 | Brand Guidelines | Design & UI | Applying your brand system to output | **160K** | | 61 | MongoDB | Data & Analytics | MongoDB and Atlas with current practice | **160K** | | 62 | Zapier | Automation | Actions across 9,000+ connected apps | **150K** | | 63 | Linear | Productivity | Planning context from Linear issues | **150K** | | 64 | Firebase | Infrastructure | Project-aware Firebase development | **150K** | | 65 | Sentry | Infrastructure | Debugging from production errors | **150K** | | 66 | Google Analytics | Data & Analytics | Read-only GA4 reporting | **150K** | | 67 | Agent Toolkit for AWS | Infrastructure | Auditable agent access to AWS | **150K** | | 68 | Airtable | Productivity | Structured records your team shares | **150K** | | 69 | SharePoint | Productivity | Governed company docs in SharePoint | **150K** | | 70 | Zoom | Communication | Meeting intelligence and Zoom apps | **150K** | | 71 | Task Master | Coding | Plans that outlive a single session | **140K** | | 72 | Brave Search | Search & Web | Broad web search from an independent index | **140K** | | 73 | Atlassian Rovo | Productivity | Jira, Confluence, and Bitbucket context | **140K** | | 74 | Perplexity | Search & Web | Sonar answers and deep research | **140K** | | 75 | Pyright LSP | Coding | Live Python type errors | **130K** | | 76 | Serena | Coding | Symbol-precise codebase navigation | **130K** | | 77 | Asana | Productivity | Team tasks and projects in Asana | **130K** | | 78 | Extract Design System | Design & UI | Design tokens from an existing site | **130K** | | 79 | Ponytail | Coding | Keeping implementations small and simple | **125K** | | 80 | Repomix | Coding | Packing a whole repo into context | **125K** | | 81 | Stripe | General | Building and debugging Stripe payments | **120K** | | 82 | Internal Comms | Writing | Status updates and announcements | **120K** | | 83 | Blender MCP | General | 3D modelling and scenes in Blender | **120K** | | 84 | Laravel Boost | Coding | Laravel work grounded in your app | **120K** | | 85 | Canvas Design | Design & UI | Posters and standalone graphics | **110K** | | 86 | DBHub | Data & Analytics | Guardrailed SQL across five engines | **110K** | | 87 | Expo | Coding | Expo and React Native app work | **100K** | | 88 | Exa | Search & Web | Agent-native web and code search | **100K** | | 89 | Neon | Data & Analytics | Branching serverless Postgres | **100K** | | 90 | Higgsfield | Media Creation | Seven-skill media generation bundle | **100K** | | 91 | ElevenLabs | Media Creation | Speech, voices, music, sound effects | **100K** | | 92 | Kubernetes MCP Server | Infrastructure | Inspecting and operating clusters | **100K** | | 93 | Graphiti | Setup & Memory | Temporal knowledge-graph memory | **100K** | | 94 | Mem0 | Setup & Memory | Portable long-term agent memory | **100K** | | 95 | Canva | Design & UI | Design production in your Canva account | **100K** | | 96 | Algorithmic Art | Media Creation | Original generative art as code | **95K** | | 97 | Calendly | Productivity | Booking links and availability | **95K** | | 98 | Matt Pocock Writing Skills | Writing | Article drafting as a guided process | **90K** | | 99 | Terraform | Infrastructure | Registry-grounded Terraform work | **90K** | | 100 | Semgrep | Coding | Security scanning while code is written | **90K** | | 101 | Prisma | Coding | Prisma ORM and Postgres operations | **90K** | | 102 | Shopify | General | Shopify store and app development | **90K** | | 103 | Granola | Communication | Decisions buried in meeting notes | **90K** | | 104 | Doc Co-Authoring | Writing | Co-writing docs section by section | **85K** | | 105 | Apify | Search & Web | Structured data from thousands of Actors | **85K** | | 106 | GitNexus | Coding | Tracing call chains and change impact | **85K** | | 107 | Redis | Data & Analytics | Live Redis data plus app patterns | **85K** | | 108 | Make | Automation | Running your existing Make scenarios | **80K** | | 109 | Miro | Productivity | Boards, diagrams, and visual context | **80K** | | 110 | Monday.com | Productivity | Boards and items in monday.com | **80K** | | 111 | ClickUp | Productivity | Tasks, Docs, and time in ClickUp | **80K** | | 112 | Hugging Face Skills | General | ML model and dataset workflows | **75K** | | 113 | Theme Factory | Design & UI | Consistent themes across artifacts | **75K** | | 114 | HubSpot | Sales | HubSpot CRM context and actions | **75K** | | 115 | Browserbase | Automation | Local or cloud browsers on demand | **75K** | | 116 | Dropbox | Productivity | Files found, summarized, and saved back | **75K** | | 117 | Otter.ai | Communication | Searching your Otter transcripts | **75K** | | 118 | Plugin Developer Toolkit | Coding | Building Claude Code plugins | **70K** | | 119 | Agent SDK Dev | Coding | Scaffolding Claude Agent SDK apps | **70K** | | 120 | Last30days | Search & Web | What people said in the last 30 days | **70K** | | 121 | Remember | Setup & Memory | Local session-to-session handoffs | **70K** | | 122 | Netlify | Infrastructure | Building and deploying on Netlify | **70K** | | 123 | DocuSign | Productivity | Agreements found, sent, and tracked | **70K** | | 124 | Fireflies | Communication | Insights from meeting transcripts | **70K** | | 125 | Power BI | Data & Analytics | DAX and Power BI semantic models | **65K** | | 126 | Lucid | Design & UI | Diagrams in your Lucid workspace | **65K** | | 127 | Salesforce | Sales | Salesforce CRM data in the agent | **65K** | | 128 | Slack GIF Creator | Media Creation | Looping GIFs that fit Slack limits | **60K** | | 129 | BigQuery | Data & Analytics | SQL on BigQuery under Google IAM | **60K** | | 130 | Greptile | Coding | Acting on Greptile review feedback | **60K** | | 131 | Learning Output Style | Setup & Memory | Learning while Claude codes | **60K** | | 132 | MCP Toolbox for Databases | Data & Analytics | One gateway to many SQL and NoSQL DBs | **60K** | | 133 | PostHog | Data & Analytics | Product analytics and feature flags | **55K** | | 134 | Webflow | Design & UI | Building and managing Webflow sites | **55K** | | 135 | Google Search Console | Marketing & SEO | Your own search performance data | **55K** | | 136 | ComfyUI | Media Creation | Node-based generation workflows | **55K** | | 137 | Google Ads | Marketing & SEO | Google Ads reporting, read-only | **50K** | | 138 | SonarQube | Coding | Your existing SonarQube quality gates | **50K** | | 139 | Jina AI | Search & Web | A wide research kit in one endpoint | **50K** | | 140 | Sourcegraph | Coding | Cross-repo search and code intelligence | **50K** | | 141 | Box | Productivity | Governed enterprise content in Box | **50K** | | 142 | Writing Guidelines | Writing | House style the agent actually follows | **50K** | | 143 | Auth0 | General | Implementing Auth0 authentication | **50K** | | 144 | Pipedream | Automation | Authenticated actions across 3,000 APIs | **50K** | | 145 | Read AI | Communication | Recaps and cross-meeting analysis | **50K** | | 146 | SEO & Content Skills | Marketing & SEO | Audits, content, and programmatic SEO | **45K** | | 147 | Humanizer ZH | Writing | De-AI-ing Chinese prose | **45K** | | 148 | Bright Data | Search & Web | Data from sites that block scrapers | **45K** | | 149 | Grafana | Infrastructure | Dashboards, metrics, and incidents | **45K** | | 150 | Snowflake | Data & Analytics | Governed Snowflake SQL and Cortex | **45K** | | 151 | Chroma | Data & Analytics | Vector store for RAG and memory | **45K** | | 152 | Interview Me | Setup & Memory | Interviewing you before it starts | **45K** | | 153 | Tableau | Data & Analytics | Governed Tableau analytics access | **45K** | | 154 | Content & Copy Skills | Marketing & SEO | Copy that converts, in every format | **40K** | | 155 | Humanizer | Writing | Stripping AI tells from English prose | **40K** | | 156 | Product Design | Design & UI | Briefs to reviewable prototypes | **40K** | | 157 | Baoyu Translate | Writing | Three-pass idiomatic translation | **40K** | | 158 | Railway | Infrastructure | Deploying and operating on Railway | **40K** | | 159 | ClickHouse | Data & Analytics | ClickHouse schemas, queries, and costs | **40K** | | 160 | OpenAI Sales | Sales | OpenAI's sales bundle for Codex | **40K** | | 161 | Datadog | Infrastructure | Production telemetry and monitors | **40K** | | 162 | DuckDB | Data & Analytics | Local DuckDB analytics guidance | **40K** | | 163 | Databricks | Data & Analytics | Governed lakehouse data and AI assets | **35K** | | 164 | Nature Skills | Writing | Academic papers and journal polish | **35K** | | 165 | YouTube Transcripts | Search & Web | Transcripts from public YouTube videos | **35K** | | 166 | PagerDuty | Infrastructure | Incident response and on-call context | **35K** | | 167 | Framer | Design & UI | Structured work on Framer sites | **35K** | | 168 | Replicate | Media Creation | Thousands of models via one MCP | **35K** | | 169 | Metabase | Data & Analytics | Governed analytics via your Metabase | **30K** | | 170 | Adobe for Creativity | Media Creation | Adobe's creative apps from Claude | **30K** | | 171 | Amplitude | Data & Analytics | Funnels, experiments, and replays | **30K** | | 172 | Ahrefs | Marketing & SEO | Ahrefs keyword and backlink data | **30K** | | 173 | Anthropic Sales | Sales | Anthropic's official sales bundle | **30K** | | 174 | Anthropic Marketing | Marketing & SEO | Anthropic's official marketing bundle | **30K** | | 175 | Lara Translate | Writing | Context-aware translation at scale | **30K** | | 176 | Intercom | General | Customer support conversations | **30K** | | 177 | Buffer | Marketing & SEO | Scheduling via your Buffer queue | **30K** | | 178 | Marketing Psychology | Marketing & SEO | Behavioral principles behind your copy | **25K** | | 179 | Stop Slop | Writing | Banning slop phrases outright | **25K** | | 180 | dbt | Data & Analytics | dbt models, tests, and MetricFlow | **25K** | | 181 | Gamma | Design & UI | Fast hosted decks from a prompt | **25K** | | 182 | HeyGen | Media Creation | Avatar videos and dubbing | **25K** | | 183 | MiniMax | Media Creation | Speech, image, video, music in one | **25K** | | 184 | Pinecone | Data & Analytics | Search and RAG on Pinecone indexes | **25K** | | 185 | Klaviyo | Marketing & SEO | Ecommerce email and SMS campaigns | **25K** | | 186 | Semrush | Marketing & SEO | Semrush competitive research | **25K** | | 187 | Apollo | Sales | Prospect search to sequence in Apollo | **25K** | | 188 | Mixpanel | Data & Analytics | Funnels, retention, and taxonomy | **25K** | | 189 | Ask Questions If Underspecified | Setup & Memory | A few must-ask questions, then go | **25K** | | 190 | Runway | Media Creation | Runway image and video generation | **25K** | | 191 | DeepL | Writing | Team translation with glossaries | **25K** | | 192 | Pulumi | Infrastructure | Infrastructure as code in real languages | **25K** | | 193 | OpenAI Image Generation | Media Creation | Images via OpenAI's native tooling | **25K** | | 194 | fal | Media Creation | 1,000+ hosted models, pay per run | **25K** | | 195 | Reddit MCP Buddy | Search & Web | Firsthand community signal from Reddit | **25K** | | 196 | Recraft | Media Creation | Production graphics and vector work | **25K** | | 197 | Hunter | Sales | Email finding and verification | **25K** | | 198 | Brevo | Marketing & SEO | Email and SMS campaigns in Brevo | **25K** | | 199 | FLUX | Media Creation | Direct FLUX.2 generation and editing | **25K** | | 200 | DaVinci Resolve MCP | Media Creation | Agent control of Resolve editing | **20K** | | 201 | OpenAI Data Analytics | Data & Analytics | Question-to-dashboard analysis flows | **20K** | | 202 | Looker | Data & Analytics | Semantic-model-aware BI on Looker | **20K** | | 203 | Public Equity Investing | General | Public-market equity research | **20K** | | 204 | SearchFit | Marketing & SEO | Free all-in-one SEO plugin | **20K** | | 205 | Attio | Sales | CRM records and pipeline in Attio | **20K** | | 206 | Slidev | Design & UI | Version-controlled developer decks | **20K** | | 207 | Metricool | Marketing & SEO | Scheduling plus social analytics | **20K** | | 208 | Pipedrive | Sales | Pipeline and deals in Pipedrive | **20K** | | 209 | Google Slides MCP | Design & UI | Editing live Google Slides decks | **20K** | | 210 | ActiveCampaign | Marketing & SEO | Lifecycle automation in ActiveCampaign | **20K** | | 211 | Substack | Marketing & SEO | Read-only Substack publication metrics | **20K** | | 212 | Crowdin | Writing | Localization projects in Crowdin | **20K** | | 213 | Hootsuite | Marketing & SEO | Enterprise social management | **20K** | | 214 | CRO Skills | Marketing & SEO | Signup, onboarding, popups, paywalls | **15K** | | 215 | Paid & Measurement Skills | Marketing & SEO | Ads, creative, and A/B testing | **15K** | | 216 | MagicPath | Design & UI | MagicPath canvas-to-code workflows | **15K** | | 217 | Lusha | Sales | Verified contacts with easy setup | **15K** | | 218 | Clay | Sales | Enrichment workflows your RevOps built | **15K** | | 219 | MotherDuck | Data & Analytics | Cloud DuckDB queries and Dives | **15K** | | 220 | Instantly | Sales | Cold-email campaigns, run live | **15K** | | 221 | ZoomInfo | Sales | Licensed B2B contact and intent data | **15K** | | 222 | Amazon Ads | Marketing & SEO | Amazon Ads campaigns and reporting | **15K** | | 223 | OpenAI Transcription | Media Creation | Transcripts with diarization guidance | **15K** | | 224 | Gong | Sales | Deal and account insights from calls | **15K** | | 225 | Circleback | Communication | Meetings, emails, and calendar context | **15K** | | 226 | tldraw | Design & UI | A shared visual canvas in Cursor | **15K** | | 227 | MailerLite | Marketing & SEO | Newsletter campaigns in MailerLite | **15K** | | 228 | Outreach | Sales | Sequences and deals in Outreach | **15K** | | 229 | Salesloft | Sales | Cadences plus Clari revenue data | **15K** | | 230 | Deepgram | Media Creation | Speech-to-text and TTS via one CLI | **15K** | | 231 | Typefully | Marketing & SEO | Drafting and scheduling X threads | **15K** | | 232 | AssemblyAI | Media Creation | Building transcription features | **15K** | | 233 | Iterable | Marketing & SEO | Iterable campaigns, safe by default | **10K** | | 234 | Paper | Design & UI | A design canvas wired to your code | **10K** | | 235 | Kit | Marketing & SEO | Creator email and newsletters on Kit | **10K** | | 236 | Picsart | Media Creation | Picsart creative API workflows | **10K** | | 237 | Customer.io | Marketing & SEO | Lifecycle messaging campaigns | **10K** | | 238 | Postiz | Marketing & SEO | Open-source posting to every network | **9K** | | 239 | Cartesia | Media Creation | Low-latency voice and TTS | **8K** | | 240 | Reply.io | Sales | Multichannel sequences in Reply | **8K** | | 241 | Smartlead | Sales | Cold-email deliverability checks | **7K** | | 242 | DataForSEO | Marketing & SEO | Raw SERP and keyword data by API | **6K** | | 243 | Close | Sales | Query and update your Close CRM | **6K** | | 244 | beehiiv | Marketing & SEO | Read-only beehiiv publication data | **6K** | | 245 | Common Room | Sales | Signal-based pipeline sourcing | **6K** | | 246 | Amplemarket | Sales | End-to-end outbound in one MCP | **6K** | | 247 | 6sense | Sales | Predictive intent and account scoring | **5K** | | 248 | Leonardo.Ai | Media Creation | Leonardo's model catalog by MCP | **5K** | | 249 | Ayrshare | Marketing & SEO | One social API across 13+ networks | **5K** |
## Plugins, skills, and MCP servers - what's the difference? Every entry in this directory is something you add to an AI agent - think of each one as an extension that gives the agent new tools, knowledge, or a working process. The same extension often ships through more than one route: as a plugin for Claude or Codex, as a generic MCP server, as a standalone skill. So we list each one once and show all of its routes, rather than listing the same thing four times under four names. * **Plugin** - a packaged bundle installed inside the agent: commands, skills, and MCP connections in one install, distributed through marketplaces. * **Skill** - a folder of instructions the agent reads and applies when relevant. The lightest route, and the most portable: the same SKILL.md format now works across many agents, with `.agents/skills/` emerging as the shared vendor-neutral location. * **MCP server** - a running service the agent connects to for live tools and data, using the open Model Context Protocol. Works in any MCP-capable agent. * **CLI** - a command-line tool the agent runs in the terminal. Works with any agent that can run shell commands. ### What each agent calls them - and how to install | Agent | What extensions are called | How you add them | | :--------------------------- | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Claude | Plugins, skills, MCP servers, CLIs. On claude.ai and Claude Desktop, hosted integrations are called connectors. | `/plugin` browses and installs from marketplaces; skill folders go in `~/.claude/skills`; `claude mcp add` registers a server. Connectors are enabled in Settings. | | Codex | Plugins, skills, MCP servers, CLIs. In ChatGPT, hosted integrations are called apps (formerly connectors). | `/plugins` installs from a marketplace repo; skills load from `.agents/skills/`; `codex mcp add` registers a server. | | Cursor | Plugins, skills, MCP servers - note that Cursor plugins are separate from classic VS Code extensions. | Install plugins from the Cursor Marketplace; skills go in `.cursor/skills/` or `.agents/skills/`; MCP servers go in `.cursor/mcp.json`. | | GitHub Copilot | MCP servers and agent skills. (GitHub App-based Copilot Extensions were sunset in November 2025 and replaced by MCP.) | MCP servers are configured in `mcp.json` in editors, or through repo Settings → Copilot on github.com, and used in agent mode; skills use the same portable skill format, in `.github/skills` or `.agents/skills`. | | Microsoft 365 Copilot | Agents and connectors. | Agents are built in Copilot Studio and distributed through Microsoft's Partner Center program; Microsoft 365 Copilot connectors bring in outside data, including MCP-based federated ones. | | Google Gemini | Gemini CLI extensions - each bundles MCP servers, commands, and context. (Not to be confused with the consumer Gemini app's Connected Apps, formerly branded "Extensions.") | `gemini extensions install `; plain MCP servers can also be configured directly in `settings.json`; skills go in `.gemini/skills` or `.agents/skills`. | | OpenCode | Plugins, skills, MCP servers. | Skills drop into `.opencode/skills/`; plugins go in `.opencode/plugins/` or install as npm packages; MCP servers are added in `opencode.json`. | | Windsurf (now Devin Desktop) | MCP servers, rules, skills. Cognition rebranded Windsurf as Devin Desktop in June 2026. | MCP servers are added through the MCP Marketplace or `~/.codeium/windsurf/mcp_config.json`; rules now prefer `.devin/rules/`, with `.windsurf/rules/` as a legacy fallback. | | Cline | MCP servers, rules, skills. | MCP servers are added by editing `mcp.json` or via the `cline mcp` CLI wizard; skills go in `.cline/skills/`. | For any agent not listed here: if it speaks MCP, add the extension's MCP server to its config; if it loads skills or instruction files, copy the skill folder into wherever it reads them from. Those two routes are the portable ones - they work the same everywhere. ## How we estimate installs No single authoritative source exists for how widely an agent extension is deployed. Vendors rarely publish telemetry and most tools ship through several channels at once, so we analyze thousands of public signals across multiple sources and reconcile them into a single estimate. **Direct installation counters receive the greatest weight.** These can include Claude plugin installs from the exact listing. skills.sh installs from the exact skill or collection page, with skill-level and collection-level counts kept separate. Visual Studio Marketplace cumulative installs, counted only where the extension materially provides the tool rather than merely mentioning it. Where these exist, they anchor the estimate and always take priority. **Distribution data helps when no authoritative counter exists.** We examine downloads for exact, verified npm and PyPI packages and pulls for Docker images that serve as a tool’s primary distribution route. Download activity is not equivalent to installed users: it can include updates, automated builds, mirrors, and repeated downloads. We therefore normalize it by channel and calibrate it against extensions for which both distribution data and direct installation counters are available. **Public implementation evidence.** We search public code for exact package names, server identifiers, and configuration entries such as those found in `.mcp.json` files. These matches can indicate that developers have implemented or evaluated a tool, but they do not prove that every configuration remains active, and they exclude usage in private repositories. **Attention metrics are used as supporting evidence.** GitHub stars, repository activity, search demand, documentation traffic, and public discussion can help validate the general order of magnitude or reveal adoption that other counters miss. They are not converted directly into installations. Repository signals are considered only when the repository is clearly associated with the tool’s primary development or distribution. **Every source is verified and overlaps are reconciled.** We tie evidence to the exact listing, package, image, or repository and exclude similarly named forks, mirrors, examples, and unrelated integrations. We also account for wrappers around existing packages, monorepos that publish the same tool through multiple channels, and traffic likely caused by continuous integration or automated systems. These estimates are intended to show relative market adoption, which we use as one directional input in our rankings: an extension with broader adoption may be better maintained, and more likely to be relevant to a larger number of users, but popularity alone does not make it the best choice. The right extension for you depends on your actual need and the tools you already run; each listing lays out capabilities, requirements, and integrations so you can make that call.