Documentation by ClonePartner

Start typing to search.

Connector Reference

Connectors overview

View Markdown

Browse every compiled connector type in registry order, see what each one does, and learn where connector configs live.

Connectors overview

A Connector is a named, stored configuration for an external system — a database path, an API token, a storage bucket. Tasks never embed these configs. A task declares connector slots (a name plus a type), and Envoy binds a stored Connector to each slot at runtime, so the same task YAML runs against any environment. See Connectors and slots for the full model.

Compiled connector types

These types are compiled into the Envoy binary and registered at startup in src/register-all.ts. Each page documents the type's configuration keys and the methods a call_method step can invoke.

Type Page Purpose
azure-blob Azure Blob Storage Read, write, and manage blobs in Azure Storage containers
csv CSV Stream records from CSV files in a local directory
yaml YAML Read and write YAML files in a local directory
sqlite SQLite Query and write a local SQLite database file
truto Truto Unified and proxy API access to SaaS apps through Truto integrated accounts
mongodb MongoDB Query, aggregate, and write MongoDB collections
dynamics365 Dynamics 365 Dataverse OData CRUD, metadata, and schema creation
postgres PostgreSQL Query and write PostgreSQL over a connection pool
mysql MySQL Query and write MySQL over a connection pool
mssql MS SQL Server Query and write Microsoft SQL Server via the tedious driver
s3 S3 and R2 Object storage on AWS S3, Cloudflare R2, and other S3-compatible stores
jira Jira Jira Cloud REST v3 with automatic pagination
freshdesk Freshdesk Freshdesk REST v2 with automatic pagination and attachment upload

The exact set available in a deployment depends on the customer build. If a task requires a type that is not registered, the run fails with Unknown connector type listing the available types.

Catalog connector types

Deployments with a connector catalog register additional types dynamically at boot — one pickable connector type per catalog integration slug (for example zendesk, hubspot, salesforce). These types are not compiled pages in this reference; they share one behavior model described in Catalog connectors and connected accounts.

Where connector configs live

  • Server and UI — Connectors are created and stored on the Connectors screen (or via the REST API) and bound to slots when you create a Job or start a Run. See Connectors and connected accounts.
  • CLI — connector configs are supplied in a connectors.yaml file passed with --config. See connectors.yaml format.

In both paths the task subprocess receives a fully resolved YAML with connector configs inlined; the pipeline does not know where the config came from.