# Dagster MCP for AI Agents AI Agent Connect

> Dagster lets you manage your entire data pipeline infrastructure through a chat interface. Instead of digging through dashboards, you can ask your agent to check job statuses, audit schedules, or find specific run logs instantly. It puts your data orchestration under your direct control without the constant tab-switching.

## Overview
- **Category:** loved-by-devs
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_9Cla5bbOzNXXdswxQ5yhHdcbDLhh8oRbE80hiC24/ai-agent-connect
- **Tags:** data-orchestration, data-pipelines, workflow-automation, observability, data-engineering, asset-management

## Description

Managing data pipelines usually feels like a constant game of whack-a-mole. You're jumping between tabs to see why a job failed, checking logs to find the error, and trying to remember which assets are actually updated. This Connector changes that by bringing your Dagster instance directly into your chat window. You can just ask what's running right now or get a summary of recent failures without ever opening a browser. It makes the way you interact with your data stack feel more like a conversation and less like a series of manual lookups. By connecting through the Vinkius marketplace, you get a direct line to your orchestration layer. It’s built for the people who need to know the health of their data right now so they can actually get back to building instead of just monitoring.

## Tools

### list_schedules
List all configured job schedules. This helps you audit your automation and timing across the project.

### list_jobs
Get a full list of all data jobs in your Dagster environment. Use this to understand your active pipeline boundaries.

### list_runs
Pull a chronological history of recent job executions. It's the fastest way to see what just finished.

### get_run
Retrieve detailed status and execution logs for a specific run ID. This is your go-to for debugging failures.

### list_assets
Enumerate all software-defined assets. Use this to identify data dependencies and storage mappings.

### list_sensors
See all active sensors listening for external events. This helps you audit your automation triggers.

## Prompt Examples

**Prompt:** 
```
Show me the status of the last 5 runs.
```

**Response:** 
```
Here are the last 5 runs for your active pipelines:

| Run ID | Job Name | Status | Completion Time |
| :--- | :--- | :--- | :--- |
| **123** | **daily_etl_job** | ✅ Success | 10:45 AM |
| **122** | **daily_etl_job** | ✅ Success | 09:45 AM |
| **121** | **ml_training** | ❌ Failed | 08:30 AM |
| **120** | **daily_etl_job** | ✅ Success | 07:45 AM |
| **119** | **sensor_sync** | ✅ Success | 07:00 AM |

Would you like to see the logs for the failed ml_training run?
```

**Prompt:** 
```
What assets are in my project?
```

**Response:** 
```
I found the following software-defined assets in your Dagster project:

*   `raw_customer_data` (Source)
*   `cleaned_sales_table` (Transformed)
*   `monthly_metrics_view` (Aggregated)

I can provide the materialization history for any of these if you need to check data freshness.
```

**Prompt:** 
```
List all jobs in my deployment.
```

**Response:** 
```
Your deployment currently has 4 active jobs:

1.  **daily_etl_job** - Handles nightly data ingestion.
2.  **ml_training_pipeline** - Trains the weekly models.
3.  **sensor_triggered_sync** - Syncs data when external files arrive.
4.  **cleanup_routine** - Removes temporary files.

Which one would you like to audit for recent runs?
```

## Capabilities

### Check active jobs
See every data job available in your deployment to understand your pipeline boundaries.

### Inspect run history
Pull up the chronological list of recent job runs to see what's been happening.

### Get specific run details
Retrieve full execution logs and statuses for a single run ID.

### Audit data assets
Enumerate all software-defined assets to see your data dependencies and storage mappings.

### Review job schedules
See all configured job schedules and sensors to audit your automation triggers.

### Verify environment status
Check your deployment boundaries and instance connectivity across clusters.

## Use Cases

### Debugging a failed pipeline
A data engineer asks the agent to show the logs for the last failed run of the daily ETL job to identify a schema mismatch.

### Auditing a new deployment
A platform team asks the agent to list all active sensors to ensure no triggers were missed during a migration.

### Checking data freshness
An analytics engineer asks what assets are defined and if the latest run was successful before updating a dashboard.

### Mapping dependencies
A developer asks the agent to list assets to see how raw customer data flows into the cleaned sales table.

## Benefits

- Stop tab-switching by checking pipeline health directly in your chat.
- Debug faster by pulling specific execution logs without hunting for the right ID in a dashboard.
- Map your data dependencies clearly by seeing how your software-defined assets connect.
- Audit your automation triggers instantly to ensure everything is firing correctly.
- Verify your deployment boundaries across clusters to ensure your instance connectivity is stable.

## How It Works

The bottom line is you get a conversational interface for your entire data orchestration layer.

1. Subscribe to this Connector and grab your Dagster URL and User API Token from your Deployment Settings.
2. Plug those credentials into your AI client to establish a secure connection to your instance.
3. Ask your agent to list jobs, check run statuses, or audit your assets in plain English.

## Frequently Asked Questions

**Can I use Dagster MCP to see if my pipelines failed?**
Yes, you can ask your agent to pull the history of recent runs to see exactly which ones failed and what their current status is.

**Does Dagster MCP work with my self-hosted instance?**
Yes, it works with both Dagster Plus and self-hosted clusters by using your unique URL and API token.

**Can I use this to find specific error logs?**
You can ask the agent to fetch the details for a specific run ID to see the execution logs and identify why a step failed.

**How do I connect my Dagster account?**
You just need your Dagster URL and a User API Token, which you can find in your Deployment Settings.

**Can I see all my data assets with this?**
Yes, you can enumerate all software-defined assets to see your dependencies and storage mappings in plain English.

**Does this help with auditing my schedules?**
It lets you list all configured job schedules and sensors to see how your automation is triggered across your project.

**Can my agent list all software-defined assets in Dagster?**
Yes. Use the 'list_assets' tool. Your agent will retrieve all software-defined assets, allowing you to identify data dependencies and verify physical storage mappings within your pipelines.

**How do I check the status of a specific job run?**
Provide the 'run_id' to the 'get_run' tool. Your agent will fetch detailed information for that specific execution, including status (Success, Failure, In Progress) and detailed execution logs.

**Can I see active sensors and schedules via the agent?**
Absolutely. Use the 'list_schedules' and 'list_sensors' tools. Your agent will pull the active automation triggers, allowing you to audit which jobs are scheduled and which sensors are listening for external events.