Skip to content
Scalekit Docs
Talk to an Engineer Dashboard

Connect AI agents to ProShort AI

Scalekit connector
Open markdown

The ProShort AI connector lets your AI agent act in each user's ProShort AI account. Each user connects their own ProShort AI API key once, and Scalekit sends it with every call, so your agent never handles credentials. It comes with 5 tools.

Tools
5
What they doRead · write · destructive
5 · 0 · 05 read0 write0 destructive
Users sign in with

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 ProShort AI connection

    In AgentKit > Connections, create a ProShort AI connection. The name you give it is the connection_name your code passes. See Configure connections.

  4. 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 = 'proshortai'
    const identifier = 'user_123'
    

    // Generate an authorization link for the user
    const { link } = await actions.getAuthorizationLink({ connectionName: connector, identifier })
    console.log('Authorize ProShort AI:', 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: 'proshortai_recordings_search',
      toolInput: {},
    })
    console.log(result)
    Terminal window
    npx tsx quickstart.mts

    Each user opens the link once and enters their ProShort AI credentials there. See Authorize a user for the full flow and statuses.

Tools

Pass the exact name to execute_tool
Try in PlaygroundRequest a tool
  • proshortai_recordings_bulk_getFetch details for up to 100 ProShort recordings in one call by their document_ids — e.g.Read-only

    Bulk Get Recordings

    Fetch details for up to 100 ProShort recordings in one call by their document_ids — e.g. to hydrate a list of results from Search Recordings. Missing/invalid IDs are reported separately in not_found rather than failing the whole request. Use projections to keep the response small when you don't need every field (especially the transcript).

    Inputs

    meeting_idsarrayrequired
    The document_ids of the recordings to fetch, 1-100 per call. Duplicates are automatically de-duplicated by the API.
    projectionsarray
    Fields to include for each recording in the response, to minimize payload size. Omit to receive all fields for every recording.
  • proshortai_recordings_getRetrieve full detail for a single ProShort meeting recording by its document_id (typically obtained from Search Recordings): title, scheduled time, platform, attendees, prospect company, AI-generated overview, transcript, and media URLs.Read-only

    Get Recording

    Retrieve full detail for a single ProShort meeting recording by its document_id (typically obtained from Search Recordings): title, scheduled time, platform, attendees, prospect company, AI-generated overview, transcript, and media URLs. Use the projections parameter to request only the fields you need and keep the response small — this matters especially for the (potentially large) transcript field.

    Inputs

    meeting_idstringrequired
    The recording's unique identifier (document_id), e.g. as returned by Search Recordings or Bulk Get Recordings.
    projectionsstring
    Comma-separated list of fields to include in the response, to minimize payload size. Omit to receive all fields. Valid field names: document_id, event_title, source, scheduled_at, meeting_platform, meeting_participants, parties, prospect_company_name, transcript, thumbnail_url, gcp_public_url, overview.
  • proshortai_recordings_get_v1Retrieve a ProShort recording via the original v1 endpoint.Read-only

    Get Recording (v1, Full Transcript)

    Retrieve a ProShort recording via the original v1 endpoint. Unlike the current v3 Get Recording tool (which returns a flattened, plain-text transcript), v1 always returns the complete raw diarized transcript — an array of speaker-attributed segments, each broken into individual words with start/end timestamps and confidence scores — plus meeting participants (with join/leave events) and parties (attendee/company details). There is no field-selection support: v1 always returns every field, with no projections parameter. If the meeting isn't available or doesn't belong to the calling customer, the response still returns HTTP 200 with an empty data object.

    Inputs

    meeting_idstringrequired
    The recording's unique identifier (document_id), e.g. as returned by Search Recordings, Bulk Get Recordings, or Get Recording.
  • proshortai_recordings_get_v2Retrieve a ProShort recording via the v2 endpoint.Read-only

    Get Recording (v2, Structured Overview)

    Retrieve a ProShort recording via the v2 endpoint. The standout difference from the current v3 Get Recording tool: when the overview projection is requested, v2 returns a rich structured object (recap, budget, timing, use_case, pain_points, action_items, prospect_info, contract_status, questions_asked, risks_obstacles, decision_process, customer_reactions, product_discussion, product_experience, product_suggestions, competitors_mentioned) instead of v3's flattened plain-text summary — useful when you need the individual deal-analysis fields rather than prose. Use projections to request only the fields you need; document_id must always be included in the projections list when projections is used. Adding projections is optional but recommended to speed up requests. If the meeting isn't available or doesn't belong to the calling customer, the response still returns HTTP 200 with an empty data object.

    Inputs

    meeting_idstringrequired
    The recording's unique identifier (document_id), e.g. as returned by Search Recordings, Bulk Get Recordings, or Get Recording.
    projectionsstring
    Comma-separated list of fields to include in the response, to minimize payload size and speed up the request. document_id must always be included in this list when projections is used. Omit entirely to receive all fields. Valid field names: document_id, event_title, source, scheduled_at, meeting_platform, meeting_participants, parties, prospect_company_name, transcript, thumbnail_url, gcp_public_url, overview.
All 26 in Transcription