Skip to content
Scalekit Docs
Talk to an Engineer Dashboard

Connect AI agents to the Qdrant Cloud MCP server

Vendor MCP
Open markdown

The Qdrant Cloud MCP connector routes your AI agent's tool calls to Qdrant Cloud's own MCP server through Scalekit. Each user signs in to Qdrant Cloud once, and Scalekit stores and refreshes their tokens, so your agent never handles credentials. It comes with 129 tools.

Tools
129
What they doRead · write · destructive
93 · 10 · 2693 read10 write26 destructive
Users sign in with
OAuth app
Your own Qdrant Cloud MCP server app

Setup

  1. Install the SDK

    Terminal window
    npm install @scalekit-sdk/node dotenv
  2. Set your credentials

    Add your Scalekit credentials to your .env file. Find values in app.scalekit.com > Developers > API Credentials.

    .env
    SCALEKIT_ENVIRONMENT_URL=<your-environment-url>
    SCALEKIT_CLIENT_ID=<your-client-id>
    SCALEKIT_CLIENT_SECRET=<your-client-secret>
  3. Create the Qdrant Cloud MCP connection

    In AgentKit > Connections, create a Qdrant Cloud MCP connection and copy its redirect URI. The name you give it is the connection_name your code passes. See Configure connections.

  4. Register an OAuth app

    Qdrant Cloud MCP server connections use your own OAuth app. Register one with Qdrant Cloud MCP server and add the redirect URI you copied.

    Then enter the app's Client ID and Client Secret on the Qdrant Cloud MCP connection.

  5. Authorize a user and make your first call

    quickstart.mts
    import { ScalekitClient } from '@scalekit-sdk/node'
    import 'dotenv/config'
    import { createInterface } from 'node:readline/promises'
    const scalekit = new ScalekitClient(
    process.env.SCALEKIT_ENVIRONMENT_URL,
    process.env.SCALEKIT_CLIENT_ID,
    process.env.SCALEKIT_CLIENT_SECRET,
    )
    const actions = scalekit.actions
    const connector = 'qdrantmcp'
    const identifier = 'user_123'
    // Generate an authorization link for the user
    const { link } = await actions.getAuthorizationLink({ connectionName: connector, identifier })
    console.log('Authorize Qdrant Cloud MCP:', link)
    const rl = createInterface({ input: process.stdin, output: process.stdout })
    await rl.question('Press Enter after authorizing...')
    rl.close()
    // Make your first call
    const result = await actions.executeTool({
    connector,
    identifier,
    toolName: 'qdrantmcp_get_authenticated_user',
    toolInput: {},
    })
    console.log(result)
    Terminal window
    npx tsx quickstart.mts

    Each user signs in once. See Authorize a user for the full flow and statuses.

Tools

Pass the exact name to execute_tool
Try in PlaygroundRequest a tool
  • qdrantmcp_doc_get_pageRead one page in full, as Markdown, including its code examples.Read-only

    Doc Get Page

    Read one page in full, as Markdown, including its code examples. For a missing page, the error suggests existing ones.

    Inputs

    pagestringrequired
    The page to read: a qdrant.tech URL, its index.md URL, or a path such as /documentation/search/hybrid-queries/, /blog/qdrant-1.15.x/, or search/hybrid-queries (a path without a source is a documentation path).
  • qdrantmcp_doc_list_docsBrowse the structure of the available content.Read-only

    Doc List Docs

    Browse the structure of the available content. Without a section, lists the top-level documentation sections and the other sources. With a section, lists its pages with a short description: documentation and the course in sidebar order, articles and the blog newest first. Long sections come in batches; the end of the output says how to get the next one.

    Inputs

    formatstring
    markdown (the default), or json for the same listing as structured data.one of markdownjson
    limitinteger
    Maximum number of pages to list.default 50
    offsetinteger
    How many pages of the section to skip.default 0
    sectionstring
    A section path such as search, /documentation/cloud/, or /blog/. Leave it out to list the top-level sections.
  • qdrantmcp_doc_search_docsSearch Qdrant content on qdrant.tech: the documentation, articles, the blog, and courses.Read-only

    Doc Search Docs

    Search Qdrant content on qdrant.tech: the documentation, articles, the blog, and courses. Each result is a page, with its source and the headings or passages on it that matched. Read the page with doc_get_page before relying on it: matches are short excerpts.

    Inputs

    querystringrequired
    What to look for, in natural language or keywords.
    formatstring
    markdown (the default), or json for the same results as structured data.one of markdownjson
    limitinteger
    Maximum number of pages to return.default 10
    partitionstring
    Optionally restrict documentation results to one part of the docs: develop (building apps and the API), deploy (installation, Cloud, operations), ecosystem (integrations and frameworks), learn (guides to search quality and production operations). Other sources ignore it.one of developdeployecosystemlearn
    sourcesarray
    Which sources to search: documentation, articles, blog, course. Leave it out to search all of them.
  • qdrantmcp_doc_search_snippetsFind code examples in Qdrant content: the documentation, articles, the blog, and courses.Read-only

    Doc Search Snippets

    Find code examples in Qdrant content: the documentation, articles, the blog, and courses. Each result is a code snippet with a description of what it does and the page it is on. Pass a language to get examples for one client. Documentation pages usually show an example in one language and link to a full snippet page with it in every client language; read that with doc_get_page.

    Inputs

    querystringrequired
    What the code should do, in natural language or keywords, such as "create a collection with sparse vectors".
    formatstring
    markdown (the default), or json for the same snippets as structured data.one of markdownjson
    languagestring
    The language of the code: python, typescript, rust, java, csharp, go, http (the REST API), shell, and so on. Leave it out to search all languages.
    limitinteger
    Maximum number of snippets to return.default 5
  • qdrantmcp_get_accountGets an account identified by the given ID.Read-only

    Get account

    Gets an account identified by the given ID.

    Inputs

    accountIdstringrequired
    No description.
  • qdrantmcp_get_account_inviteGets an account invite identified by the given account ID and invite ID.Read-only

    Get account invite

    Gets an account invite identified by the given account ID and invite ID.

    Inputs

    accountIdstringrequired
    No description.
    inviteIdstringrequired
    No description.
  • qdrantmcp_get_account_memberGets an account member by ID.Read-only

    Get account member

    Gets an account member by ID. The authenticated actor must be a member of the same account as the member being fetch.

    Inputs

    accountIdstringrequired
    No description.
    userIdstringrequired
    No description.
  • qdrantmcp_get_account_quotasGet quotas for the account identified by the given account ID.Read-only

    Get account quotas

    Get quotas for the account identified by the given account ID.

    Inputs

    accountIdstringrequired
    No description.