Skip to content

Repster

Lightweight scheduling and observability for dlt pipelines.

dlt loads your data; Repster runs it. Add one decorator to an existing dlt pipeline and get a scheduler, run history, and a web UI, all behind an AI-ready CLI that does everything, so your coding agent can drive it too.

import dlt
import repster as rp # add this

@rp.flow() # add this
def flow_name():
    """ Your dlt pipeline code goes here """
    ...

uv add repster

How it works

Illustrative example: an hourly schedule around a dlt pipeline:

import dlt
import repster as rp

@rp.flow(schedule="0 * * * *")   # run hourly
def github_stars():
    @dlt.resource
    def stars():
        yield from fetch_stars("my-org/my-repo")

    pipeline = dlt.pipeline(destination="duckdb")
    pipeline.run(stars())

Run it locally:

rp run github_stars       # run now
rp dev                    # start the scheduler and local UI

What you get

Everything here is open source (Apache 2.0) and runs entirely on your machine:

  • An AI ready CLI. Everything is an rp command, from running flows and viewing logs to deploying, managing secrets, and more. The web UI is an alternative view, never the only way.

  • One decorator. @rp.flow() wraps your pipeline while leaving dlt fully usable underneath. Its state, destinations, and tooling all work as before.

  • Just run rp dev. Repster schedules your flows, records every run, and serves a web UI where you can trigger flows, watch logs, and inspect run history.

Wondering how this compares to cron, Airflow, or Airbyte? See Why Repster?

Because every operation is available from the CLI, AI coding agents can drive the entire development loop. Write a flow, run it, inspect the logs, fix the code, retry the run, and deploy, all without needing a browser.

Repster also ships with an agent skill, and these docs are published as llms.txt for agent context windows. See Use Repster with AI coding agents.

Repster Cloud

When pipelines need to run without your laptop open, deploy them:

rp deploy

Then flows can run on a schedule or on-demand:

rp run --cloud github_stars

Repster Cloud gives you:

  • Serverless compute: each run gets its own isolated worker, created on demand and gone when the run finishes. Nothing to provision, size, or keep running. Automatically scales to your workload, and you only pay for what you use.
  • Managed scheduling: the same cron expressions from your code, fired by the cloud.
  • Teams and secrets: authentication, org roles, and an encrypted secrets vault.
  • Transparent pricing: when metered billing launches, you'll pay for raw compute: CPU seconds, memory, storage. No seats, no rows, no MTU.

Develop locally, then deploy with confidence. The same flow code you test on your machine runs in Repster Cloud, so moving from development to production is seamless.

Deploy to cloud →

Get started

Quickstart →: build and run your first flow in 5 minutes

Deploy to cloud →: move from local to managed execution

Sections

Section What you'll find
Tutorials Step-by-step guides for getting started
How-to guides Recipes for specific tasks
Reference CLI commands, configuration, and API
Explanation Concepts and background