Agent Reference
Open source · Apache 2.0. This section documents the open-source core. The Data Workers platform is covered separately - see What the platform adds.
Nine agents ship in the open-source core. Each is an independent MCP server: enable the ones you want, ignore the rest. None depends on another to function.
| Agent | Start with | Does |
|---|---|---|
| Pipelines | dw-pipelines | Pipeline generation, validation, deployment planning |
| Incidents | dw-incidents | Anomaly detection and root-cause analysis |
| Context & Catalog | dw-context-catalog | Cross-catalog search, lineage, asset classification |
| Quality | dw-quality | Profiling, quality scoring, anomaly detection |
| Schema | dw-schema | Schema diffing, evolution, compatibility checks |
| Governance | dw-governance | Compliance checks, classification, policy |
| Observability | dw-observability | Metrics, SLA monitoring, pipeline tracing |
| Connectors | dw-connectors | Catalog connector management and discovery |
| Usage Intelligence | dw-usage-intelligence | Usage analysis, query tracking, index recommendations |
Two further agents - MLOps and Orchestration - are in the repository and not yet wired into the CLI.
Ask your client to list tools for any agent to see its current surface. The tool list is the reference; we don't maintain a second copy of it here that can drift.
Which to start with
Incidents and Context & Catalog. Both answer questions you'd otherwise chase by hand, and neither changes anything.
Where the open-source core stops. The open-source agents read, analyse, and recommend. They run against a built-in sample estate so you can drive the full tool surface without wiring anything up. Pointing them at your own infrastructure, letting them act on what they find, and the browser console are the Data Workers platform. → What the platform adds · Book a demo