Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.mantrixflow.com/llms.txt

Use this file to discover all available pages before exploring further.

​
Dashboard

After signing in, the Dashboard (/workspace) is the landing page. It shows:
  • Active Pipelines — count of running pipelines out of total
  • This Week — number of migrations run
  • Rows Processed — total rows from recent runs
  • Team Members — active members in the organization
  • Usage & Limits — plan quota for Pipelines, Rows This Month, API Calls, and Agent Interactions with upgrade prompts

​
Organizations

An organization is the workspace boundary for connections, pipelines, team members, activity, and schedules. Many teams keep one organization for production and one for staging so credentials and run history stay separate.

​
Connections

A connection is a reusable credential set plus connector-specific configuration. Connections live under Connections and are created with a role of Source or Destination. Examples in the current product:
  • a PostgreSQL source connection for an application database
  • a Shopify source connection for store data
  • a GitHub source connection for one repository
  • a PostgreSQL, MySQL, or SQLite destination connection for downstream tables

​
Pipelines

A pipeline begins as a lightweight shell (name + source connection) created from Data Pipelines → + New Pipeline. Clicking Create & open canvas opens the canvas builder with a Source node on the left and a Destination node on the right.
  • Click ⚙️ on the Source node to select tables, discover schema, and preview rows.
  • Click ⚙️ on the Destination node to open the five-tab Destination panel:
    • Config — connection, final delivery schema, sync mode, write mode, validate
    • Normalisation — column normalisation settings
    • dbt Layer — dbt model integration
    • Preview — target table preview
    • Scheduling — recurring schedule configuration

​
Branches

A branch is a connection line on the canvas from the Source node to a Destination node. One source can have multiple Destination nodes. A realistic pattern:
  • Destination A writes cleaned orders to PostgreSQL for analytics.
  • Destination B writes a narrower version to MySQL for an internal view.

​
Sync modes

The app currently supports two read patterns:
  • Full sync — available now; recommended for all first runs and production pipelines
  • Incremental sync — beta; for tables with a dependable cursor field such as updated_at; not fully production-tested

​
Destination behavior

The Destination Config tab has three write mode options:
  • Upsert — available now; inserts new rows and updates existing rows by primary key
  • Append — coming soon; insert-only, no deduplication
  • Replace — coming soon; drops and recreates the destination table on each run
Use Upsert for all production workflows today.

​
Activity Log

The Activity Log (/workspace/activity) shows all events across pipelines, data sources, and the organization — with columns for Timestamp, Entity Type, Action Type, Message, and Status. Filter by type, action, status, or time window.

​
Run history and sync state

Every run stores operational metadata: status, timing, row counts, and errors. Access per-pipeline run history from the history icon in the builder top bar. Incremental pipelines also maintain sync state so later runs know where to resume.

​
Mental model to keep in mind

An organization owns connections, a pipeline shell points to a source connection, the builder turns that shell into one or more branches, and run history tells you whether the configured flow actually worked.