Skip to content
Scalekit Docs
Talk to an Engineer Dashboard

Connect AI agents to the Fuse AI MCP server

Vendor MCP
Open markdown

The Fuse AI MCP connector routes your AI agent's tool calls to Fuse AI's own MCP server through Scalekit. Each user signs in to Fuse AI once, and Scalekit stores and refreshes their tokens, so your agent never handles credentials. It comes with 305 tools.

Tools
305
What they doRead · write · destructive
114 · 12 · 179114 read12 write179 destructive
Users sign in with
OAuth app
Your own Fuse AI 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 Fuse AI MCP connection

    In AgentKit > Connections, create a Fuse AI 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

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

    Then enter the app's Client ID and Client Secret on the Fuse AI 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 = 'fuseaimcp'
    const identifier = 'user_123'
    // Generate an authorization link for the user
    const { link } = await actions.getAuthorizationLink({ connectionName: connector, identifier })
    console.log('Authorize Fuse AI 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: 'fuseaimcp_contact_call_engagement_summary',
    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
  • fuseaimcp_add_spintax_to_text_toolRewrites a piece of email/message text with spintax {variation|variation} alternatives to improve deliverability.Read-only

    Add Spintax To Text

    Rewrites a piece of email/message text with spintax {variation|variation} alternatives to improve deliverability. Returns the spintax-enriched text without saving it; if the AI output is unusable or would alter merge fields, placeholders or existing spintax, the original text is returned unchanged.

    Inputs

    bodyTextstringrequired
    The email/message text to enrich with spintax
  • fuseaimcp_build_prospect_search_filtersUse it first when the user asks FuseAI to find people or companies — find/see ('find me...', 'show me...'), 'create a list of...', 'add to my list...', and compound discovery+campaign requests.Read-only

    Build Prospect Search Filters

    Use it first when the user asks FuseAI to find people or companies — find/see ('find me...', 'show me...'), 'create a list of...', 'add to my list...', and compound discovery+campaign requests. Not for specific people the user names or that were already found ('add Steve, the CRO at Datadog, to my list', 'save these 3'): save those with save_people_to_list, which finds each one's real profile itself. Converts the request into Prospect Search filters, looking up the company names, job titles, places and similar values in the request with Fuse's external people and company data providers so the filters use values the search recognizes. Whenever any filter can be built, you must then run the search with them and present the first page of results. Criteria that cannot be expressed as filters are left out of the filters and reported back, with what to tell the user and what to offer. Creates NOTHING — no list, no saved contacts, no campaign — and charges no credits.

    Inputs

    entityTypestringrequired
    Which Prospect Search tab the filters are for. "people" when the target is humans (job titles, roles, seniority). "companies" when the target is organizations (industries, headcount, funding, locations).one of peoplecompanies
    querystringrequired
    Natural language description of the people or companies to find. Include EVERY targeting criterion the user gave: job titles, seniority, departments, industries, company size, revenue, funding, location, skills, named companies, etc. Do NOT include a requested result count — counts are ignored for prepared searches.
  • fuseaimcp_check_company_savedCheck which companies the user has saved in their CRM, by website domain.Read-only

    Check Saved Companies

    Check which companies the user has saved in their CRM, by website domain. Pass `domains` (1 to 500 company domains or website URLs; each is normalized to its registrable domain). Returns { data } with one row per input domain: { domain, isSaved, companyId? } where isSaved is true only when this user has saved the company, and companyId is the CRM company id whenever a company record exists for that domain (it can exist without being saved by this user). Use it before creating a company or when a flow needs to know whether a company is already in the user's CRM.

    Inputs

    domainsarrayrequired
    No description.
  • fuseaimcp_check_contact_has_emailCheck whether the user holds a usable email for a contact saved in FuseAI, based on the email statuses already stored (no new verification or other outside call is made).Read-only

    Check Contact Has Email

    Check whether the user holds a usable email for a contact saved in FuseAI, based on the email statuses already stored (no new verification or other outside call is made). Pass contactId (24-hex) and optional verifiedOnly (default true: only verified-valid emails count; false: any deliverable email counts, including unverified user-supplied ones). Returns { hasEmail, email } where email is the primary email when it qualifies, else the first qualifying email, else null. Use it as a gate before an email step.

    Inputs

    contactIdstringrequired
    No description.
    verifiedOnlyboolean
    No description.default true
  • fuseaimcp_check_linkedin_relationCheck whether any LinkedIn account the user owns, or one a teammate shared with them, is connected to a LinkedIn profile (by its slug/public id), and list pending invitations sent to or received from that person.Read-only

    Check LinkedIn Relation

    Check whether any LinkedIn account the user owns, or one a teammate shared with them, is connected to a LinkedIn profile (by its slug/public id), and list pending invitations sent to or received from that person. Returns isRelation, the connected accountIds and those invitations. It reads FuseAI's synced copy and does not contact LinkedIn, so a very recent change may not show yet.

    Inputs

    relationIdstringrequired
    The LinkedIn profile slug / public id (the part after /in/ in the profile URL, e.g. 'john-doe-123')
  • fuseaimcp_check_sequence_list_compatibility_toolPre-flight check before attaching a list to a sequenced campaign: verifies that at least one contact in the list has a valid email and/or a LinkedIn URL, as the sequence's channels require (when both are required, one contact must have both; at least one of isEmailRequired / isLinkedInRequired must be true).Read-only

    Check Sequence/List Compatibility

    Pre-flight check before attaching a list to a sequenced campaign: verifies that at least one contact in the list has a valid email and/or a LinkedIn URL, as the sequence's channels require (when both are required, one contact must have both; at least one of isEmailRequired / isLinkedInRequired must be true). Returns whether it is compatible (and why not). Use it when unsure a list can actually be sent the planned sequence.

    Inputs

    isEmailRequiredbooleanrequired
    Whether the sequence contains email steps
    isLinkedInRequiredbooleanrequired
    Whether the sequence contains LinkedIn steps
    listIdstringrequired
    The prospect list _id (24-character Mongo id) to check
  • fuseaimcp_contact_call_engagement_summaryPer-contact power-dialer engagement digest over a time range (default: the 30 days up to timeRangeTo, or up to now; read-only): total call attempts, connected conversations, per-disposition counts, latest outcome with note and a short transcript snippet with explicit speaker labels.Read-only

    Contact Call Engagement Summary

    Per-contact power-dialer engagement digest over a time range (default: the 30 days up to timeRangeTo, or up to now; read-only): total call attempts, connected conversations, per-disposition counts, latest outcome with note and a short transcript snippet with explicit speaker labels. Use to decide which leads to double down on or drop — e.g. 'connected twice but no next step' or 'tried 4+ times, never answered'. Optional: timeRangeFrom/timeRangeTo (ISO dates), sortBy (most-connected default | most-attempted | most-recent), minimumTotalCalls, maximumConnectedCalls (set to 0 for never-connected leads), latestDispositions, limit (default 20, max 50), page. For a specific contact's transcripts, follow up with get_contact_call_transcripts.

    Inputs

    latestDispositionsarray
    No description.
    limitinteger
    No description.default 20
    maximumConnectedCallsinteger
    No description.
    minimumTotalCallsinteger
    No description.
    pageinteger
    No description.default 1
    sortBystring
    No description.one of most-connectedmost-attemptedmost-recentdefault most-connected
    timeRangeFromstring
    No description.
    timeRangeTostring
    No description.
  • fuseaimcp_count_all_enriched_contacts_toolReturns how many contacts are in the dynamic 'All Enriched Contacts' list.Read-only

    Count All Enriched Contacts

    Returns how many contacts are in the dynamic 'All Enriched Contacts' list.

    Inputs

    This tool takes no inputs.