dltHub
Book a demo

dltHub AI harness

The dltHub AI harness: guardrails your coding agent can't skip

Your interface to dltHub from local development. The harness equips Claude Code, Codex, or Cursor with 10 toolkits of skills, rules, and MCP tools covering the whole lifecycle: ingest, transform, explore, and operate. Your agent follows workflows dltHub maintains instead of improvising, on infrastructure dltHub runs.

Works withClaude CodeCodexCursor
Cl

Your coding agent

Claude Code, Codex, or Cursor

4 specialized subagents

Ingestion

REST, SQL, and file sources

Modeling

transformations and the CDM

Analytics

notebooks and dashboards

Operations

deploy, run, fix, and maintain

operates through
dltHubthe agentic layer

dltHub AI harness

Skills, commands, rules, and MCP: the agentic primitives to build, run, and fix pipelines

you are here

dltHub context catalog

Lineage, schema, data quality, governance, run state, and API context assets

grounds & runs

Sources

  • REST APIs

  • SQL databases

  • Files on S3, GCS, Azure

dltHub

dltHub

the managed infrastructure layer

  • Ingestion & transformation

  • Scheduling & runs

  • Data quality & lineage

Warehouse

  • Snowflake

    Snowflake

  • Databricks

    Databricks

  • BigQuery

    BigQuery

  • DuckDB

    DuckDB

Serving

  • Notebooks & dashboards

  • Data apps

  • Agents & MCP clients

dltHub — agentic data layer + managed infrastructureYour agent & external tools

Start with one prompt

For builders

Build your first dlt pipeline

Paste this prompt into Claude, Codex, or Cursor. The agent does the rest.

Run uvx dlthub-start@latest to build my first pipeline and run it on dltHub

The dltHub AI harness: your team owns Day 2

Your interface to dltHub from local development. The harness equips your coding agent (Claude, Codex, Cursor) with the skills, platform tools, and run context to operate ingestion and transformation pipelines in production as a team: run, fix, maintain, and then build new ones to your conventions. Your data team and their agents own Day 2, no DevOps or platform engineering required.

Cheatsheet|ai-harness.md
Start
Quick Startdlt1 skill · 1 cmd · 1 rule · MCP
Initdlt3 skills · 1 cmd · 1 rule · MCP
Ingest
REST API Pipelinedlt8 skills · 1 rule · MCP
SQL Database Pipelinedlt8 skills · 1 rule · MCP
Filesystem Pipelinedlt3 skills · 1 rule · MCP
Harden
Data QualitydltHub4 skills · 2 rules · MCP
Performancedlt1 skill · 1 rule · MCP
Transform
TransformationsdltHub6 skills · 1 rule · MCP
Explore
Data Explorationdlt2 skills · 1 rule
Operate
dltHub PlatformdltHub4 skills · 2 rules

Every skill in the harness

What your agent reaches for, step by step, from first prompt through the pipelines it keeps running.

Quick Start1/1

The guided entry point. Names a use case, checks the workspace, and hands off to the right toolkit in a few prompts.

Opus 5.0 · Quick Start · ~/pipelines

>
? for shortcuts
Ask your agentcopies with harness setup

Take me through the full workflow with the GitHub API