Noloco

@Noloco list all active customer records from the Customers table, sorted by signup date, and show their email and subscription tier.

Retrieved 47 active customer records from your Noloco app's Customers table. The list includes email addresses and subscription tiers (Free: 12, Pro: 28, Enterprise: 7), sorted by signup date from January 2024 to present. The most recent signups are three Enterprise accounts from last week. This example shows the first page of results; pagination cursors are available if you need to retrieve additional records beyond the initial batch.

This example demonstrates read-access to custom Noloco tables. The MCP must call NOLOCO_DESCRIBE_TABLE first to learn your app's exact field names and available filters—field schemas vary by account. Requires API key with read permissions for the target table.

@Noloco create a new project record with title 'Q2 Marketing Campaign', status 'Planning', and assign it to user ID 'usr_abc123'.

Created project record #PRJ-1847 in your Noloco Projects table with the specified attributes. The record is now visible in your Noloco app with status 'Planning' and assigned to the user. The MCP validated field types against your app's schema before submission—for instance, confirming that 'status' accepts 'Planning' as a valid enum value and that the user ID exists. This write operation is immediate and cannot be undone through the MCP.

Write operations require the MCP to first inspect your table schema via NOLOCO_DESCRIBE_TABLE to ensure field names and types match. This example shows a creation side-effect: the record persists in your Noloco database immediately. API key must have write permissions for the target table.

@Noloco pull all support tickets marked 'urgent' from the last 7 days and draft a summary email for the team highlighting common themes.

Retrieved 14 urgent support tickets from your Noloco Tickets table (filtered by priority='urgent' and created_at >= 7 days ago). Common themes identified: 8 tickets involve payment processing errors, 4 relate to mobile app login failures, and 2 concern data export timeouts. Draft email: 'Team—urgent ticket volume is up 40% this week. Payment gateway issues dominate (57% of cases), likely tied to the API migration on March 3rd. Mobile auth and export performance also need immediate attention.' This synthesis pairs Noloco's filtered data retrieval with AI reasoning to surface actionable patterns.

This example showcases the MCP's ability to filter records by date and priority, then feed structured data into the AI for synthesis. The quality of the summary depends on ticket descriptions being present in your Noloco schema. Rate limits apply to bulk reads; large result sets may require pagination.

When field teams need a write-back layer without dev time

A 6-person field service crew logs job completions, equipment failures, and customer sign-offs into a Noloco app their ops manager built in two days. The crew lead uses Switchy to parse photos of handwritten notes, extract structured data, and write records directly into Noloco tables via the MCP—no Zapier delay, no CSV upload ritual. The MCP's describe-then-create pattern means the AI learns the exact field names and types on the fly, so schema changes don't break the workflow. This works when your Noloco app is the system of record and updates happen under 50 per day; beyond that, you'll want batch endpoints Noloco doesn't expose yet. If your field data lives in Airtable or a proper CRM, skip this MCP and use those native integrations instead.

Noloco as a read-only knowledge base for support agents

A 3-person support team at a B2B SaaS company uses Noloco to host a client directory with contract terms, feature flags, and support tier metadata. Agents ask Switchy "what's Acme Corp's renewal date" and the MCP fetches the record in under two seconds—faster than clicking through the Noloco UI and hunting for the right table. The list-tables and describe-table tools let the AI navigate unfamiliar schemas without a human writing a lookup script. This breaks down when your Noloco app has more than 20 tables or complex relational queries; the MCP doesn't support joins, so you'll end up chaining multiple get-record calls and burning tokens. If your client data is already in HubSpot or Salesforce, use those MCPs—they handle relationships natively.

When you need conditional record updates during live events

A 4-person nonprofit runs weekend workshops and tracks RSVPs, attendance, and no-shows in a Noloco app. During check-in, a volunteer asks Switchy to mark attendees as "checked in" or flag dietary restrictions that weren't captured at signup. The MCP's update-record tool changes only the fields the volunteer mentions, leaving the rest untouched—critical when you're working with incomplete voice input. The delete-record tool handles duplicate signups on the spot. This setup works when updates are ad hoc and under 100 per event; if you're processing hundreds of check-ins, you need a barcode scanner and a proper event platform, not an AI middleman. Noloco's API key auth means the volunteer's Switchy session has full write access, so this only works when your team trusts each other with destructive operations.