Skip to content
Scalekit Docs
Talk to an Engineer Dashboard

Connect AI agents to the TheirStack MCP server

Vendor MCP
Open markdown

The TheirStack MCP connector routes your AI agent's tool calls to TheirStack's own MCP server through Scalekit. Each user signs in to TheirStack once, and Scalekit stores and refreshes their tokens, so your agent never handles credentials. It comes with 79 tools.

Tools
79
What they doRead · write · destructive
52 · 10 · 1752 read10 write17 destructive
Users sign in with
OAuth app
Your own TheirStack 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 TheirStack MCP connection

    In AgentKit > Connections, create a TheirStack 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

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

    Then enter the app's Client ID and Client Secret on the TheirStack 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 = 'theirstackmcp'
    const identifier = 'user_123'
    // Generate an authorization link for the user
    const { link } = await actions.getAuthorizationLink({ connectionName: connector, identifier })
    console.log('Authorize TheirStack 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: 'theirstackmcp_count_companies_by_country',
    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
  • theirstackmcp_count_companies_by_countryList every ISO country code with the number of companies headquartered in that country, for geographic distribution analysis.Read-only

    Count companies by country

    List every ISO country code with the number of companies headquartered in that country, for geographic distribution analysis. Free: it doesn't consume credits.

    Inputs

    countrystring
    Name of a country. Will return the country matching their name in English
    limitinteger
    Number of results per pagedefault 25
    pageinteger
    Page number.default 0
  • theirstackmcp_count_jobs_by_countryList every country code with the number of jobs located in that country and the number of companies hiring there.Read-only

    Count jobs by country

    List every country code with the number of jobs located in that country and the number of companies hiring there. Search by country name in English. Free: it doesn't consume credits.

    Inputs

    countrystring
    Name of a country. Will return the country matching their name in English
    limitinteger
    Number of results per pagedefault 25
    pageinteger
    Page number.default 0
  • theirstackmcp_count_webhook_eventsCount the events a webhook has triggered, optionally only those with a given delivery status.Read-only

    Count webhook events

    Count the events a webhook has triggered, optionally only those with a given delivery status. Use it when the number is all you need, such as how many deliveries failed.

    Inputs

    webhook_idintegerrequired
    ID of the webhook
    statusstring
    Status of the webhook events to filter byone of IN_PROGRESSSUCCESSFAILEDNOT_TRIGGERED
  • theirstackmcp_count_webhook_events_per_dayCount the team's webhook events per day over a date range (the last 30 days by default), with the credits they consumed, for one webhook or all of them.Read-only

    Count webhook events per day

    Count the team's webhook events per day over a date range (the last 30 days by default), with the credits they consumed, for one webhook or all of them. Optionally split each day by delivery status, webhook or event type, e.g. to chart delivery health over time.

    Inputs

    end_datetimestring
    End of the range, in ISO format (e.g. 2023-01-01T23:59:59). Defaults to now.
    group_byarray
    Dimensions to split each day's count by: delivery status, webhook or event type. Several can be combined.
    start_datetimestring
    Start of the range, in ISO format (e.g. 2023-01-01T00:00:00). Defaults to 30 days before the end.
    timezonestring
    IANA timezone string (e.g., 'Europe/Madrid', 'America/New_York')default UTC
    webhook_idinteger
    Webhook ID
  • theirstackmcp_estimate_saved_search_volumeReturns how many records a saved search matched per day over the last 7 days, and what they would cost.Read-only

    Estimate a saved search's daily volume

    Returns how many records a saved search matched per day over the last 7 days, and what they would cost. Use it to size max_events_per_day: a cap below avg_records_per_day means the webhook will drop matches every day, and dropped matches are never sent later. A webhook whose events are companies (a company search, or trigger_once_per_company) has no daily estimate and returns nulls with an unavailable_reason.

    Inputs

    search_idintegerrequired
    ID of the saved search to estimate
    trigger_once_per_companyboolean
    Whether the webhook fires once per company. A webhook that does has no daily estimate, because its events are companies rather than jobs and it only fires for companies it has not sent yet.default false
  • theirstackmcp_find_locationsSearch for geographic locations by name or ID to retrieve standardized location data including city, region, and country.Read-only

    Find locations

    Search for geographic locations by name or ID to retrieve standardized location data including city, region, and country. Use the location IDs in the job_location_or and job_location_not filters of a job search to filter by city or region.

    Inputs

    country_codestring
    The country code of the location, all the returned locations will be filtered by this country code
    feature_codearray
    GeoNames feature code(s) to filter by location type (e.g. PPL=populated place, ADM1=first-order admin division). See https://www.geonames.org/export/codes.html
    idarray
    The ids of the locations
    limitinteger
    The limit of the resultsdefault 5
    namestring
    Search value for the location, this will be used to filter the locations by name
    offsetinteger
    The offset of the resultsdefault 0
  • theirstackmcp_find_technologies_and_topicsFind technologies and buying intent topics in TheirStack's keyword catalog, and the slugs the technology and keyword filters of the search tools take.Read-only

    Find technologies and topics

    Find technologies and buying intent topics in TheirStack's keyword catalog, and the slugs the technology and keyword filters of the search tools take. Search by name with `q`; results are sorted by how many companies have each keyword. Pass `keyword_type=technology` for technologies only, or `keyword_type_not=technology` for buying intent topics only. There are about 42k keywords, so search rather than page through them. Free: it doesn't consume credits.

    Inputs

    category_patternstring
    Case-insensitive text to search for in the keyword category. For example, `data` to find 'data tools'.
    category_slugstring
    Keyword category slug
    include_metadataboolean
    Deprecated, metadata is always included nowdefault false
    keyword_typestring
    Keyword type to filter byone of technologysoftware_producttechnology_conceptoperational_activitystrategic_initiativemetric_or_conceptproduct_or_service_offeredregulationthreat_or_riskeventcommodityindustry_verticalphysical_equipmentmedical_researchgovernment_body
    keyword_type_notarray
    Exclude keywords of these types
    limitinteger
    Number of results per pagedefault 25
    name_patternstring
    Case-insensitive text to search for in the keyword name
    pageinteger
    Page number.default 0
    parent_category_patternstring
    Case-insensitive text to search for in the keyword parent category name
    parent_category_slugstring
    Keyword parent category slug
    qstring
    General search query. Searches across keyword name, keyword type, category, and parent category (case-insensitive). Results where the keyword name starts with the query are ranked first.
    slugarray
    Keyword slug of a single keyword
    slugsarray
    Keyword slugs of one or many keywords
    starts_withstring
    Filter keywords whose name starts with this string (e.g. 'A' for glossary). Special value '__other__' selects keywords whose name does not start with letters A-Z.
  • theirstackmcp_generate_app_urlBuild a deep link into the TheirStack app for a given resource and return it as `{"url": ...}`.Read-only

    Generate app URL

    Build a deep link into the TheirStack app for a given resource and return it as `{"url": ...}`. Pick the resource with the `type` field: `company_list` (view a list), `job_search` / `company_search` (open a search from the same filters a job or company search takes), `saved_search` (open a saved search), or `webhook` (view a webhook). Search links (`job_search` / `company_search`) are workspace-independent — the encoded query opens against whoever's session loads it, so the link is portable when shared. The other types point at a workspace-specific resource and include the workspace, which is taken from the authenticated caller, so it is never passed in.

    Inputs

    filtersobject
    The same filters a job search takes (`search_jobs`, `POST /v1/jobs/search`). The URL opens the jobs search with these filters applied and runs the search automatically.
    list_idinteger
    ID of the company list to open
    saved_search_idinteger
    ID of the saved search to open
    search_typestring
    Whether the saved search is a jobs or a companies searchone of jobscompanies
    typestring
    Selects the variant. Accepted values: 'company_list', 'job_search', 'company_search', 'saved_search', 'webhook'. 'company_list' uses type, list_id; 'job_search' uses type, filters; 'company_search' uses type, filters; 'saved_search' uses type, saved_search_id, search_type; 'webhook' uses type, webhook_id. Send only the fields belonging to the selected variant.one of company_listjob_searchcompany_searchsaved_searchwebhook
    webhook_idinteger
    ID of the webhook to open