twenty_skill

Overview

This skill provides an open-source CRM API for managing people, companies, opportunities, notes, tasks and custom objects via REST or GraphQL. It exposes endpoints for CRUD operations, metadata discovery, and webhooks so you can integrate CRM data into other systems or build custom workflows. Setup requires an account and an API key or a self-hosted URL.

How this skill works

The skill uses REST endpoints for core CRUD actions (companies, people, tasks, notes, opportunities) and a GraphQL endpoint for complex relational queries. Metadata endpoints reveal object schemas and picklists for custom fields. Authentication is handled with a Bearer API key passed in the Authorization header and requests can be paginated using limit/offset or cursor-based pageInfo.

When to use it

• Sync contacts, companies or deals between systems
• Build or extend integrations that need CRM CRUD operations
• Query object schemas and custom fields before mapping data
• Automate task and note creation from external triggers
• Listen and react to real-time events via webhooks

Best practices

• Store the TWENTY_API_KEY and TWENTY_API_URL in environment variables
• Use GraphQL for complex joins and REST for simple CRUD calls
• Use metadata endpoints to discover custom object fields before writing data
• Respect cloud rate limits; consider batching or backoff on 429 responses
• Wrap curl calls that use environment variables inside bash -c to avoid variable clearing when piping

Example use cases

• Create a sync job that imports leads from a form into people and companies endpoints
• Build a dashboard that lists opportunities and their related company data using GraphQL
• Automatically create follow-up tasks and notes when a deal changes stage
• Expose picklist values from metadata to drive consistent dropdowns in a UI
• Subscribe to webhook events to trigger downstream automation in real time