📚 Product & AI Agent Documentation
Explore how toorow unifies marketing data, enforces semantic governance, and exposes governed FastMCP tools for AI agents.Introduction
Product overview, key capabilities, and strategic direction.
Supported Connectors
Explore the 37 pre-built connector modules (GA4, Meta, Shopify, Stripe, etc.).
MCP Host Integration
Connect external LLM host clients (Claude Desktop, Cursor, Antigravity).
FastMCP Agent Tools
Exhaustive API reference for FastMCP data tools exposed to AI agents.
MCP App Use Cases
Interactive workflows rendering embedded React/TypeScript UI widgets.
Agent Context & Metadata
Dynamic schema discovery and business context event annotations.
Universal Datastreams
Real-data ingestion pipelines across APIs, BigQuery, and managed feeds.
File-Source Ingestion
Ingesting media plans, audience panels, and TGI extracts.
Semantic Layer & Marts
Canonical daily fact surface (
fact_daily_kpi) and dbt ratio views.Anomaly Detection
Surveillance methodology, baseline stats, and z-score alert thresholds.
Agentic Daily Insights
Automated morning briefings and narrative insights synthesis.
Agent Rules
System instructions and behavioral invariants for LLMs.
🛠️ Self-Hosted & Infrastructure (Tab: Self-Hosted)
Deploy and operate toorow on your own infrastructure or cloud environment.Quickstart
Install and run toorow locally in under 5 minutes.
Prerequisites
Cloud accounts, GCP infrastructure, and local dev tooling.
Self-Hosting Guide
Production deployment steps and complete
.env variable reference.Cloud Hosting
Maintainers’ production GCP Cloud Run architecture and scale-to-zero settings.
Platform Constraints
Security checklists, zero-telemetry rules, and architectural invariants.
Adding a Connector
Scaffold a new connector module with step-by-step Mailgun walkthrough.
Repository Boundary
Public application projection vs. private workspace rules.