AI_Calling Retell HubSpot MCP Server
Enables AI assistants to check tour availability, book tours via HubSpot Scheduler, and log call sessions to Supabase.
README
AI_Calling — Retell HubSpot MCP Server
Middle layer between Retell AI (Cara), HubSpot Scheduler/CRM, and Supabase for tour availability, booking, and call logging.
Retell call → Render MCP server → HubSpot Scheduler API + CRM → Supabase logging
Architecture
| MCP tool | Purpose |
|---|---|
get_tour_availability |
Check HubSpot availability for virtual or in-person tours |
book_tour |
Book a tour after guest confirms |
log_retell_session |
Save Retell session metadata to Supabase |
log_tour_preference |
Log tour interest without booking |
HTTP routes
| Route | Auth | Purpose |
|---|---|---|
GET /health |
No | Render health check |
POST /mcp |
Bearer | MCP Streamable HTTP endpoint for Retell |
POST /webhooks/retell |
Bearer | Retell post-call webhook → Supabase |
POST /cron/sync-call-data |
Bearer | HubSpot deals (ai_call_attempted=true) → Retell calls → Supabase |
Local setup
- Copy environment file:
cp .env.example .env - Fill in
.envwith your real values (never commit.env). - Run Supabase SQL from
supabase/schema.sqlin the Supabase SQL editor. - Install and start:
npm install npm run dev - Verify health:
curl http://localhost:3000/health
Environment variables
Use these exact names locally (.env) and in Render (Dashboard → Web Service → Environment).
| Variable | Description |
|---|---|
PORT |
Server port (3000 locally; Render sets automatically) |
NODE_ENV |
development or production |
MCP_SERVER_SECRET |
Random secret; Retell sends Authorization: Bearer <this> |
HUBSPOT_ACCESS_TOKEN |
HubSpot Private App Retell Connection token only |
HUBSPOT_API_BASE |
https://api.hubapi.com |
SUPABASE_URL |
Supabase project URL |
SUPABASE_SERVICE_ROLE_KEY |
Supabase service_role key (server-side only) |
DEFAULT_TIMEZONE |
Europe/Madrid |
HUBSPOT_IN_PERSON_MEETING_URL |
In-person meeting page URL |
HUBSPOT_VIRTUAL_MEETING_URL |
Virtual meeting page URL |
HUBSPOT_IN_PERSON_SLUG |
info-madrid |
HUBSPOT_VIRTUAL_SLUG |
info-madrid/virtual-tour-booking-carabanchel |
DEFAULT_TOUR_DURATION_MINUTES |
30 |
RETELL_API_KEY |
Retell API key (required for call sync) |
RETELL_API_BASE |
https://api.retellai.com |
HUBSPOT_AI_CALL_ATTEMPTED_PROPERTY |
ai_call_attempted |
HUBSPOT_RETELL_CALL_ID_PROPERTIES |
Comma-separated deal properties that store a Retell call id |
SYNC_ENABLED |
true to run incremental sync on an interval inside the web service |
SYNC_INTERVAL_MS |
3600000 (1 hour) when SYNC_ENABLED=true |
SYNC_INITIAL_DELAY_MS |
60000 — delay before the first incremental run after server start |
Where to get secrets
- HubSpot token: Settings → Integrations → Private Apps → Retell Connection → Access token
- Supabase service role: Supabase → Project Settings → API →
service_rolekey - MCP secret: Generate a long random string; use the same value in Retell MCP headers
Important: Do not use the Supabase publishable (anon) key for server-side inserts if RLS is enabled. Use the service role key in Render only — never expose it client-side or in Retell.
Important: Do not use the old Carabanchel_Direct_Booking_Integration HubSpot app. Use Retell Connection only.
HubSpot scopes required
The Retell Connection private app needs:
crm.objects.contacts.writecrm.objects.contacts.readcrm.objects.deals.writecrm.objects.deals.readcrm.schemas.deals.readcrm.objects.contacts.read(contact phone lookup during call sync)crm.schemas.contacts.readcrm.objects.owners.readautomationscheduler.meetings.meeting-link.readmeetings-write← required for booking; add before live booking tests
Render deployment
| Setting | Value |
|---|---|
| Build command | npm ci --include=dev && npm run build |
| Start command | npm start |
| Health check path | /health |
| Instance | Starter recommended (Free tier cold starts can exceed Retell's 20s MCP timeout) |
Do not use yarn or yarn start — this repo uses npm (package-lock.json). Using yarn alone skips the TypeScript build and can resolve incompatible dependency versions.
Recommended Render settings:
| Setting | Value |
|---|---|
| Build command | npm ci --include=dev && npm run build |
| Start command | npm start |
| Health check path | /health |
If Render still uses yarn for build, the app will start via tsx (commit 1a6921c+) but you should switch to npm for reliable installs. Node 22 is required (see .nvmrc).
Add all environment variables from the table above in Render → Environment.
After changing build/start commands in the Render dashboard, trigger Manual Deploy → Clear build cache & deploy.
Connect Retell MCP
In Retell → MCP → Add MCP:
| Field | Value |
|---|---|
| MCP server URL | https://ai-calling-j1hu.onrender.com/mcp |
| Timeout | 20000 |
| Headers | Authorization: Bearer <MCP_SERVER_SECRET> |
Select these tools after deploy:
get_tour_availabilitybook_tour(requires HubSpotmeetings-writescope)log_retell_sessionlog_tour_preference
Optional webhook (post-call): POST https://ai-calling-j1hu.onrender.com/webhooks/retell with the same bearer header.
Cara prompt behavior
When the guest wants a tour:
- Ask whether they prefer a virtual or in-person tour.
- Ask which day works best.
- Ask what Madrid time generally works best.
- Call
get_tour_availability. - Offer one or two available slots.
- Ask clearly: "Should I book that for you?"
- Only after the guest clearly confirms, call
book_tour. - Do not say the tour is booked until
book_tourreturns success. - If booking fails, say: "No problem, I'll send you the tour links by WhatsApp so you can choose the time that works best for you."
- Log the result with
log_retell_sessionorlog_tour_preference.
Manual test scripts
# Test HubSpot availability (requires HUBSPOT_ACCESS_TOKEN in .env)
npx tsx scripts/testAvailability.ts
# Test Supabase inserts (requires SUPABASE_* in .env)
npx tsx scripts/testSupabaseInsert.ts
# Test live booking (guarded — only runs when enabled)
RUN_BOOKING_TEST=true TEST_BOOKING_EMAIL=you@example.com npx tsx scripts/testBookTour.ts
Pre-live checklist
- Run
supabase/schema.sqlin Supabase - Deploy to Render with all env vars
- Confirm
GET /healthreturns{ "ok": true } - Run
testAvailability.tsandtestSupabaseInsert.ts - Connect Retell MCP with bearer auth
- Test a call with your own HubSpot contact and phone
- Add
meetings-writescope if booking fails with missing scope errors - Only then enable for real leads
Supabase tables
retell_sessions— Retell call/session metadata plus HubSpot deal/contact enrichment (upsert bysession_id)mcp_tool_calls— MCP tool request/response logstour_bookings— Tour bookings and preferencessync_runs— Audit log for HubSpot deal → Retell call sync jobs
See supabase/schema.sql for full schema. Existing projects should also run:
HubSpot fields stored on retell_sessions after sync:
| Column | Source |
|---|---|
hubspot_deal_id |
Deal ID |
hubspot_deal_name |
Deal name |
hubspot_pipeline |
Pipeline label (e.g. Hubs B2C - Carabanchel) |
hubspot_deal_stage |
Stage label (e.g. Contacted) |
hubspot_deal_stage_id |
Stage internal ID |
hubspot_unit_type |
unit_type__carabanchel_ |
hubspot_contract_start_date |
desired_check_in_date |
hubspot_contract_end_date |
desired_check_out_date |
hubspot_contact_id |
Associated contact ID |
hubspot_contact_name |
Contact first + last name |
hubspot_contact_email |
Contact email |
hubspot_contact_phone |
Contact phone |
HubSpot deal → Retell call sync
Syncs deals where ai_call_attempted = true into retell_sessions, linked by hubspot_deal_id.
Matching order (first hit wins):
- Deal property in
HUBSPOT_RETELL_CALL_ID_PROPERTIES(default:retell_call_id,retell_session_id,ai_retell_call_id) - Retell call
metadata.hubspot_deal_id - All syncable Retell calls to the deal's associated contact phone (connected and failed dials:
dial_no_answer,dial_busy,telephony_provider_permission_denied)
Failed attempts are stored with end_reason = the Retell disconnection reason and session_outcome = failed_dial.
One-time historical backfill (local or Render shell):
npm run backfill:calls
Ongoing sync — option A: Render Cron Job (recommended)
| Setting | Value |
|---|---|
| Schedule | 0 * * * * (hourly) |
| URL | POST https://ai-calling-j1hu.onrender.com/cron/sync-call-data |
| Header | Authorization: Bearer <MCP_SERVER_SECRET> |
Ongoing sync — Option B: in-process scheduler (active in render.yaml)
Set on the Render web service:
| Variable | Value |
|---|---|
SYNC_ENABLED |
true |
SYNC_INTERVAL_MS |
3600000 |
SYNC_INITIAL_DELAY_MS |
60000 |
RETELL_API_KEY |
Your Retell API key |
The server runs an incremental sync ~60s after startup, then every hour. Only deals modified since the last successful sync are re-processed (24h overlap buffer). Overlapping runs are skipped if a previous sync is still in progress.
Redeploy after changing environment variables in Render (Manual Deploy if needed).
Manual full re-sync:
curl -X POST "https://ai-calling-j1hu.onrender.com/cron/sync-call-data?full=true" \
-H "Authorization: Bearer <MCP_SERVER_SECRET>"
Security
- No secrets in code or
.env.example .envis gitignored- Bearer auth on
/mcpand/webhooks/retell - Logs scrub Authorization headers and token values
- All tool inputs validated with Zod
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.