Momentum
Rising startups →- Launches
- Growth
- Product
- Hiring & team
- Money
- News & stories
- Launches All MCPs launched on PeerPush
- Launches All MCPs launched on Product Hunt
- Launches All MCPs launched on PeerPush
- Launches All MCPs on Show HN
- Launches All MCPs launched on Launch Llama
Astracia Signal Score
Compare with similar →Not enough public signals for a score yet. Missing signals are not a sign of weakness.
Open source Open source alternatives to All MCPs →
Connections
Sign in to followNo public connections on record yet.
About the startup
AllMCPs is the open directory and registry for the Model Context Protocol (MCP) ecosystem — the definitive place to discover, evaluate, and install MCP servers for Claude, Cursor, Windsurf, and any MCP-compatible AI agent. As the number of MCP servers has exploded across GitHub, npm, and independent registries, developers have lost a single trustworthy place to search by capability, compare install options, and verify that a server actually does what it claims. AllMCPs solves that with a fast, semantically searchable catalog, one-click configuration generation, and a public API built specifically so AI agents — not just humans — can query, evaluate, and submit listings programmatically. At its core, AllMCPs indexes MCP servers across categories like databases, file systems, browser automation, web search, developer tooling, productivity, and communications. Every listing includes structured metadata: the tools it exposes, its install method (npx, uvx, Docker, remote HTTP/SSE), required environment variables, links to source and documentation, and community signals like popularity and verification status. Search isn't just keyword matching — it's ranked by relevance to natural-language queries like "read and write files locally" or "query a Postgres database," so both humans and LLM agents can find the right server without knowing its exact name in advance. A standout feature is the Config Generator: pick a server and a client (Claude Desktop, Cursor, etc.), and AllMCPs produces the exact JSON snippet to drhttps://allmcps.com/api/mcp and interact with the directory using the same protocol it's cataloging — a genuinely dogfooded, self-referential design. Through that endpoint, an agent can search the directory, pull a ready-to-use install config for any listed server, list categories with counts, and — critically — submit a brand-new MCP server for indexing entirely on its own, without a human filling out a web form. Submission includes an ownership-verification flow: maintainers can prove control of a listing via a GitHub README badge, a website badge, or a DNS TXT record, which unlocks a verified badge and lets them keep listing details current. For maintainers who want a first-party integration rather than going through the remote endpoint, AllMCPs also publishes allmcps-server, a small local MCP server (installable via npx allmcps-server) that exposes a single focused submit_mcp tool — point your agent at a GitHub repo or website URL and it handles the submission for you, no browser required. AllMCPs is free and open — there's no paywall on search, discovery, or basic listings. Optional paid boosts (featured placement, category sponsorship, priority review) exist for maintainers who want extra visibility, but the core registry is built to stay a neutral, comprehensive public good for the MCP ecosystem, the same way package registries and app stores serve their respective ecosystems. As MCP adoption grows across coding assistants, browser agents, and autonomous workflows, AllMCPs aims to be the layer that keeps discovery fast, installation friction-free, and trust verifiable — for the humans building agents and the agents doing the building themselves. Built by Jackalope Digital, AllMCPs i AllMCPs is free and open — there's no paywall on search, discovery, or basic listings. Optional paid boosts (featured placement, category sponsorship, priority review) exist for maintainers who want extra visibility, but the core registry is built to stay a neutral, comprehensive public good for the MCP ecosystem, the same way package registries and app stores serve their respective ecosystems. As MCP adoption grows across coding assistants, browser agents, and autonomous workflows, AllMCPs aims to be the layer that keeps discovery fast, installation friction-free, and trust verifiable — for the humans building agents and the agents doing the building themselves.
Alternatives to All MCPs
Browse category →Reviewed matches are labeled. Other suggestions come from a shared subcategory, use case, source listing or the same category. Compare features and current plans before choosing.
| Product |
|---|
RunAgentSecured, reliable AI agent deployment & integration at scale |
DXT - Find Awesome MCP ServersDXT is leading AI Agent platform. Browse 15,000+ MCP servers |
JinaAI Quality Engineer |
AgentX 2.0Create your custom cross-platform multi-agent AI team |
PDF RedactionSecure your PDFs with AI |
Botlance.aiEasy builder for AI chatbots and Agents |
Bind AIBest alternative to Lovable, Cursor and Windsurf |
A2A ProtocolEmpowering AI Agents to Collaborate Seamlessly |
More in AI & Automation
Browse category →Selected by category and available use-case details. A topic match does not establish equivalent features.
| Product |
|---|
Viston AITurning enterprise data into practical, scalable AI outcomes |
PromptyxAI Prompt Management & Engineering Platform |
WaltonWalton is a friendly AI for cowork and agents. |
visual-ai-agentsVisual ai agents builder that gives production ready code |
PanesA private feed for your AI agents and tools |
openagentemailUnlimited agent mailboxes on your own domain. |
Recently added in AI & Automation
Browse category →Added to this directory in the past 30 days. Dates refer to the directory listing, not the product launch.
| Product |
|---|
With AliaOne AI companion with memory, continuity & RU/EN support |
Unforgettable Car AIMake the card-swipe car reveal video from one photo, pay once, no subscription. |
Ghostface AITurn one selfie into grainy late-90s horror photos for the viral Ghostface AI trend. |
AI Dancing DogTurn one photo of your dog into a dancing TikTok video. |
Explore expertise in AI & Automation
See all →Profiles grouped by topic. This does not confirm an offer, availability or a connection to this profile.
| Profile |
|---|
Jorge Ibáñez "Champe"Spanish content creator, author and digital consultant focused on gaming and creator economy and Artificial Intelligence |
dragosrAutonomous Carbon Based LLM with 42 years of tuning on Information Attack and Defense. |
johnSonder.bsky.socialA Cabbage of a Man. |
Media about AI & Automation
See all →Shows and channels grouped by topic. A recommendation does not mean this profile was featured or endorsed.
| Show |
|---|
Exploring Language ModelsNewsletter · Maarten Grootendorst |
EpisodesNewsletter · Emily St. James |
Hit Points by Nathan BrownNewsletter · Nathan Brown |
Founder stories from AI & Automation startups
See all →Retold founder accounts from startups in the same category, each linked to the original.
| Story |
|---|
Wade Foster charged $100 after building the integration a forum asked forWade Foster says Zapier's first paying customer was Mixergy's Andrew Warner, billed $100 into an unfinished beta after Wade found a stale PayPal-to-Highrise plea and then shipped Wufoo-to-AWeber instead, taking the money |
FounderPal sold strategies as one-time payments, not subscriptionsSveta Bay says Marketing Strategy Generator brought around $10,000 a month in one-time payments after a second MVP earned $60,000 in six months. Dan Kulkov later said that paid product made around $100,000 in ten months. |
Prompt Genie shipped on a Friday and called 30 days five figures, not MRRKushank Aggarwal says a Carrd-and-PickAxe ChatGPT prompt generator went from idea to first paid user in four days and to five-figure revenue in 30 days via organic TikTok and Instagram, then slowed with wrapper-like chur |
Integromat's 2019 review counted users, not a $107 million saleIntegromat's February 2020 year-in-review, the assigned company original for Ondrej Gazda's Make story, reports 77,000 users at the end of 2018 and 205,000 at the end of 2019, with paying users tripling; those are adopti |
Browse AI & Automation categories
See all →Subcategories in the same part of the directory, with the number of listed products.
| Category |
|---|
AI assistants & chatbots3,871 listed products |
Machine learning platforms540 listed products |
AI models & APIs474 listed products |
Agent builders & orchestration435 listed products |
Workflow automation & integrations393 listed products |
LLMOps & evaluation189 listed products |
