Starburst

Data & Analytics

Neotask on OpenClaw makes your Starburst data lakehouse conversational — running read-only SQL queries and engaging Starburst's data product AI agents through a single business interface.

What You Can Do

The Starburst integration gives Neotask 2 data query actions.

  • `queryReadOnly` — execute read-only SQL (SELECT, SHOW, EXPLAIN) against your Starburst cluster with configurable result size limits. Returns JSON with columns and row data.
  • `data_product_agent` — engage a stateful AI agent configured within a Starburst data product for domain-specific analytical conversations
  • Every action runs autonomously or requires your approval — you decide.

    Try Asking

  • "Query Starburst for total revenue by region for Q1, joined across our data lake and warehouse"
  • "Show me all available schemas and tables in our Starburst Galaxy cluster"
  • "Explain how the query plan for our daily reporting query looks — is it optimized?"
  • "Engage the sales analytics data product agent and ask it about pipeline velocity trends"
  • Pro Tips

  • Use `queryReadOnly` for operational reporting and ad hoc analysis — the read-only constraint makes it safe to delegate to agents without risk of data modification
  • `data_product_agent` is powerful for domain experts: each data product has its own AI with deep context about that specific business area
  • Schedule daily data quality checks: your agent runs SHOW TABLE STATUS queries across critical datasets and alerts your team to anomalies
  • Combine Starburst queries with your reporting tools to automatically populate dashboards and deliver scheduled data briefings to business stakeholders
  • Works Well With