Findabis
This server has a remotely hosted endpoint. Open its connector listing for connection details, authentication requirements, and setup instructions:
What is this?
Findabis is a directory of UK medical cannabis clinics, pharmacies and products.
This MCP server lets AI assistants like Claude, ChatGPT and Gemini look things up in it directly, so answers about clinics, prices and products come from current data rather than guesswork.
It's hosted, so there's nothing to install or run.
Related MCP server: AU-BizIntel MCP Server
How it works
You ask your assistant a question. It picks one of the Findabis tools, calls it over the open Model Context Protocol, and answers from what comes back, with a link to the Findabis page behind every fact. Your chat never goes to Findabis: only the tool call does, the same request the website itself makes.
Getting started
Add this URL to your AI assistant:
https://findabis.com/api/mcpThat's it. No account, no API key.
Claude (web, desktop and mobile)
Open Settings → Connectors and choose Add custom connector.
Name it Findabis and paste the URL above. Leave the OAuth fields empty: no login is needed.
In a new chat, open the tools menu, turn Findabis on, and ask your question.
Claude Code
claude mcp add --transport http findabis https://findabis.com/api/mcpCursor
One click: it opens Cursor and adds a server named findabis with the URL above. Cursor needs to be installed on the machine you click from. Or paste the config from the section below.
Use this button rather than the one on cursor.directory for now. That listing installs the server under the nameserver and passes it a type field from the plugin manifest that Cursor's own config doesn't use.
Gemini CLI
gemini extensions install https://github.com/Findabis/MCPChatGPT
Open Settings → Apps & Connectors. Custom servers sit under Advanced settings (developer mode).
Create a connector named Findabis, paste the URL above, and choose No authentication.
Add Findabis to a chat from the + menu and ask your question.
VS Code / GitHub Copilot — add .vscode/mcp.json to your project (or run MCP: Add Server from the command palette), start the server from the file, then pick Findabis in Copilot Chat's tools list in agent mode:
{
"servers": {
"findabis": { "type": "http", "url": "https://findabis.com/api/mcp" }
}
}Cursor by hand (~/.cursor/mcp.json):
{
"mcpServers": {
"findabis": { "url": "https://findabis.com/api/mcp" }
}
}Windsurf (~/.codeium/windsurf/mcp_config.json):
{
"mcpServers": {
"findabis": { "serverUrl": "https://findabis.com/api/mcp" }
}
}Gemini CLI by hand (~/.gemini/settings.json, then restart; /mcp lists the tools):
{
"mcpServers": {
"findabis": { "httpUrl": "https://findabis.com/api/mcp" }
}
}Anything else that speaks MCP: streamable HTTP at the URL above (POST). A plain GET on the same URL returns a description of the server; the registry card lives at
/.well-known/mcp/server.json, and there's a copy in server.json.
Check it's working
Ask: "How many clinics are listed on Findabis, and what does a first consultation typically cost?"
Your assistant should call get_directory_facts and answer with counts, a fee range and links to the directory. If it answers from memory instead, Findabis isn't switched on in that chat yet (see Troubleshooting).
What an answer looks like
Every answer links to the Findabis page it came from, so you can check a figure yourself. The server also reminds the assistant, on every connection, that Findabis is a directory and not medical advice.
Things you can ask
"What's Blue Dream, and what does Findabis list for it?"
"Which clinics near Leeds do video consultations, and what's the first appointment?"
"Compare these three clinics on what a new patient pays in the first year."
"What's actually in this oil — carrier oil, allergens, THC and CBD?"
"Find me pharmacies near SW1A that deliver."
"How many clinics are on Findabis, and what does a first consultation typically cost?"
"Is this pharmacy registered, and what did its last GPhC inspection say?"
The tools
There are ten, and they all just read:
Tool | What it's for |
| Clinics that can prescribe, nearest first from a postcode or town, with fees, consultation types, CQC rating and reviews |
| Pharmacies that dispense a prescription you already have, with delivery and collection |
| Products by name, brand, strain or form, with THC/CBD and the latest published UK price |
| Everything public about one product: strength, terpenes, ingredients, allergens, leaflets |
| Strain profiles by name or genetics |
| The brands behind UK products, by name or dose form, with how many products each lists |
| One clinic, pharmacy, strain or brand page in full, with its FAQs and opening hours |
| Two to six clinics or pharmacies side by side |
| One search across the lot, plus Findabis articles, when you're not sure what you're looking for; every hit comes with a one-line summary |
| How many clinics, pharmacies, products, strains and brands are listed, the spread of first-consultation fees, how many pharmacies deliver |
The list tools return up to 25 at a time; pass offset for the next page (each answer says where it starts).
There are also three prompts, which show up as slash-commands in Claude Desktop, Claude Code and Cursor: find_care_near (a town or postcode), product_facts (a product name) and check_provider (a clinic or pharmacy name). And three guides the assistant can read: how to get a prescription in the UK, who clinics take, and what Findabis is.
What it can and can't see
It only sees what's already public on findabis.com. No accounts, no patients, no personal data — ever.
It can't book, buy, change or send anything.
It doesn't share stock levels, which pharmacy holds what, or per-pharmacy medicine prices. Those stay with patients and professionals who are signed in.
Please be gentle with it: 60 requests a minute per client is plenty for normal use. Bulk-copying the directory isn't allowed under the Findabis Terms.
The important bit
Findabis is a directory, not a clinic or a pharmacy, and nothing here is medical advice. It tells you what providers offer, never what anyone should take. That's a conversation for you and a specialist doctor.
In the UK, cannabis-based medicines are prescription only, and only a specialist can prescribe them. Having cannabis without a prescription is still an offence.
The server tells your AI assistant all of this at the start of every session, so it should pass it on too.
Troubleshooting
What you see | What it means |
The assistant answers without using Findabis | Findabis isn't switched on in that chat. In Claude, open the tools menu and turn it on; in Cursor and Gemini CLI, restart after adding the server. |
| 60 requests a minute from one address. A shared office or VPN address shares that budget. Wait a minute and ask again. |
| The client opened an old-style SSE connection. The server speaks streamable HTTP, so choose HTTP or streamable HTTP when adding it, or update the client. |
| The address must be exactly |
A page link redirects to coming soon | Findabis hasn't opened to the public yet. The tools already read the live directory; the pages open at launch. |
Found a problem?
If it's a security issue, please email info@findabis.com rather than opening a public issue — there's more in SECURITY.md. For anything else (wrong data, a missing clinic, ideas), issues are very welcome.
Source code
The code behind the public server is in src/, so you can see exactly what each tool does and what it returns.
It runs inside the Findabis app and uses the app's own database and helpers, so it won't run on its own from this repo. It's here so it can be read and checked. Findabis's internal tools aren't included.
Where | What |
The endpoint: rate limits and the connection | |
Registers the tools with their read-only hints, the guides and the prompts | |
The three prompts and what each asks the assistant to do | |
The tool-routing guide every session gets after the disclaimer | |
The ten tools and the shape of what they return | |
The medical disclaimer every session gets |
The diagrams in this README are SVG files in assets/, generated by scripts/make-docs-images.py so they can be redrawn when the tools change. The assistant marks are their owners', used only to name the products that connect.
Releasing a new version, and how it reaches the MCP Registry and the other directories, is in RELEASING.md.
Licence
The files in this repo are MIT. The directory data the server returns isn't — that's covered by the Findabis Terms. And the Findabis name and logo are ours, so please don't use them as your own.
Made in the UK by Findabis · say hi to the cat 🐾 Findabis Ltd · Company no. 16806797 · ICO ZC265580 · Changelog
This server cannot be deployed
Maintenance
Related MCP Connectors
Search, compare and find alternatives across a directory of AI agents, MCP servers and skills.
Find MCP servers, skills, plugins, agents, prompts and rules for any task.
Live AI-native web search with citations. One tool for every MCP client. Flat per-request pricing.
Pay-per-use tool marketplace for AI agents. Search, price-check, and call APIs via MCP.
Related MCP Servers
- FlicenseNot gradedqualityNot gradedmaintenanceEnables users to search and discover over 500 AI agent endpoints and x402 APIs directly from MCP-compatible clients like Claude and Cursor. It provides tools to query endpoints by keyword, category, or capability while offering access to detailed provider statistics.-
- AlicenseNot gradedqualityDmaintenanceGives any MCP-compatible AI agent instant access to the Australian Business Register (ABR) — plus AI-powered business intelligence. Search 8M+ registered Australian entities by name or ABN, get full profiles, check GST status, and get an AI-generated opportunity assessment for any business.38 npmMIT
- AlicenseNot gradedqualityDmaintenanceAn MCP server that lets AI assistants search the UK Register of Licensed Visa Sponsors (125,000+ companies), enabling queries about company sponsorship, location, visa routes, and ratings.1MIT
- AlicenseNot gradedqualityDmaintenanceThis MCP server provides authoritative medical data from FDA, WHO, PubMed, RxNorm, and Google Scholar, enabling drug information lookup, health statistics retrieval, medical literature searching, and clinical guidelines access directly in AI workflows.97 npmMIT