Dagster Slack

0.29.0 · active · verified Sat Apr 11

The `dagster-slack` library provides a Slack client resource for integrating Slack notifications and messaging capabilities into Dagster data pipelines. It enables users to send messages to Slack channels from ops, assets, schedules, and sensors, facilitating real-time communication about pipeline status and events. The current version is 0.29.0, and it follows the frequent release cadence of the broader Dagster ecosystem.

Warnings

gotcha When interacting with the Slack client directly from the `SlackResource`, use `slack.get_client().chat_postMessage(...)` instead of `slack.chat_postMessage(...)`. The `get_client()` method ensures you are calling the method on the underlying `slack_sdk.WebClient` instance.
Fix: Always call `.get_client()` on the `SlackResource` instance before invoking Slack API methods.
breaking When configuring hooks and sensors (e.g., `make_slack_on_run_failure_sensor`), the `dagit_base_url` parameter has been deprecated and replaced with `webserver_base_url` for generating deep links to runs in the UI.
Fix: Update `dagit_base_url` parameters to `webserver_base_url` in your sensor and hook definitions.
gotcha To use `dagster-slack`, you must first create a Slack App in your workspace, configure bot users, install the app to your workspace, and obtain a bot token. This token should be stored securely, ideally as an environment variable (e.g., `MY_SLACK_TOKEN`), and passed to the `SlackResource`.
Fix: Follow the official Slack App creation and token generation guide to set up your integration.
deprecated In `make_slack_on_run_failure_sensor`, the `job_selection` and `monitor_all_repositories` parameters are deprecated. Use `monitored_jobs` for specifying jobs to monitor.
Fix: Migrate from `job_selection` to `monitored_jobs` in `make_slack_on_run_failure_sensor` definitions.

Install

pip install dagster-slack Install dagster-slack

Imports

SlackResource
```
from dagster_slack import SlackResource
```

make_slack_on_run_failure_sensor

from dagster_slack import make_slack_on_run_failure_sensor

make_slack_on_step_failure_hook

from dagster_slack import make_slack_on_step_failure_hook

Quickstart

This quickstart demonstrates how to define a `SlackResource` with a Slack bot token (loaded from an environment variable) and then use it within a Dagster asset to post a message to a specified Slack channel. A Slack App and bot token are required for this to function.

import os
from dagster import Definitions, EnvVar, asset
from dagster_slack import SlackResource

@asset
def send_slack_message(slack: SlackResource):
    """An asset that sends a message to Slack."""
    channel = os.environ.get('SLACK_CHANNEL', '#general') # Ensure channel exists
    slack.get_client().chat_postMessage(channel=channel, text=":wave: Hello from Dagster!")

defs = Definitions(
    assets=[send_slack_message],
    resources={
        "slack": SlackResource(token=EnvVar("MY_SLACK_TOKEN"))
    },
)

view raw JSON →