# Coalesce MCP for AI Agents AI Agent Connect > Coalesce MCP lets your AI agent manage Snowflake data pipelines, trigger jobs, and monitor execution logs directly. It connects your agent to the Coalesce API so you can handle complex data transformations through chat instead of clicking through a complex UI. Use it to check environment configurations, debug failed runs, or kick off new transformation jobs on demand with simple natural language commands. ## Overview - **Category:** developer-tools - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_Cc4llz27o6QgNOF3cmyYyEYcNpp6loEG8VfVdFq3/ai-agent-connect - **Tags:** data-transformation, snowflake, etl-pipeline, data-engineering, job-monitoring, column-aware ## Description Imagine you're in the middle of a sprint and a production pipeline stalls. Instead of hunting through the Coalesce dashboard to find the right environment and digging through logs to see what broke, you just ask your agent. This Connector gives your AI a direct line to your Coalesce organization. It can pull environment configs, check the status of the last run, and even kick off new transformation jobs on demand. It's built for the reality of data engineering where you need to move fast without constant context switching. By adding this to your toolset via Vinkius, you turn your AI into a command center for your Snowflake workflows. You can debug failed runs or check the health of your dev environment with a simple question, letting you stay focused on the actual logic rather than the overhead of managing the infrastructure. You can quickly check if a run is still materializing data in Snowflake or if it has finished, providing instant clarity on your data readiness. This means no more tab switching or manual ID hunting. You get a direct way to manage your data transformation lifecycle through natural language, making it easier to keep your data moving without the friction of a complex UI. ## Tools ### trigger_job Start a specific job in a target environment immediately. This lets you kick off a transformation without leaving your chat or editor. ### trigger_run Start a new run for an entire environment or a specific job. It is the quickest way to refresh your data pipelines on demand. ### get_environment Get the full configuration details for a specific Coalesce environment. This helps you see the exact settings and IDs for your various workspaces. ### get_job_details Pull specific information about a single job within your organization. Use this to see the history and configuration of a specific transformation task. ### get_run_status Check how far along a triggered run is and what its current status is. This gives you real-time visibility into your active data pipelines. ### list_environments See every environment configured in your Coalesce organization at once. It is a fast way to get an overview of your entire infrastructure. ### list_jobs Get a list of all jobs, with the option to filter by a specific environment. This helps you find the right transformation task in a large organization. ### list_nodes Retrieve metadata for all transformation nodes in a chosen environment. Use this to inspect the components of your data pipelines. ## Prompt Examples **Prompt:** ``` Show me all my Coalesce environments. ``` **Response:** ``` You have 3 environments: * **Production** (env-001) — Last run: success, 2h ago * **Staging** (env-002) — Last run: success, 5h ago * **Development** (env-003) — Last run: failed, 1h ago Would you like to check the failed run details for the Development environment? ``` **Prompt:** ``` Trigger the 'daily_sync' job in the staging environment. ``` **Response:** ``` Done. I've triggered the **'daily_sync'** job in the **staging** environment. **Run ID:** run-8k3m **Status:** running **Estimated completion:** ~4 minutes ``` **Prompt:** ``` What's the status of the production pipeline? ``` **Response:** ``` The production pipeline (Run ID: run-9g1l) is currently at **85%** completion. * **Dimension tables:** Refreshed successfully * **Fact nodes:** Currently materializing in Snowflake ``` ## Capabilities ### Trigger transformation jobs Start specific pipelines in your environments on demand via chat. ### Monitor job status Get real-time updates on whether a run is succeeding or failing. ### Inspect environment configs View the settings and configurations of all your Coalesce environments. ### View execution logs Read the detailed logs of previous runs to debug errors quickly. ### List transformation nodes See the metadata for transformation nodes within a specific environment. ## Use Cases ### Production Failure Debugging A production pipeline fails in the middle of the night. You ask your agent to check the status of the last run and pull the logs to find the error immediately. ### On-Demand Data Refresh You need to refresh a specific set of tables for a stakeholder. You tell the agent to trigger a specific job in the staging environment to verify the data. ### Environment Audit You're onboarding a new team member. You ask the agent to list all environments and their configurations to get them up to speed on the current setup. ### Real-time Status Check You're in a meeting and need to know if the morning load finished. You ask the agent for the status of the production run to give an instant update. ## Benefits - Stop manual dashboard hunting by using list_environments to see your entire organization setup in one chat. This lets you quickly identify the correct environment ID without having to log into the web portal and click through multiple menus. - Debug failed runs much faster by asking your agent to pull logs using get_job_details when a pipeline stalls. You can see the specific error messages immediately, which saves you from digging through complex execution histories manually. - Execute tasks instantly with trigger_job to kick off specific pipelines in your staging or production environments. You can initiate these runs directly from your chat interface, which keeps you in your flow and away from the browser. - Keep a pulse on production by using get_run_status to see real-time progress of active pipelines. You can check if a run is still materializing data in Snowflake or if it has finished, providing instant clarity on your data readiness. - Audit your infrastructure by using list_nodes to inspect the metadata of all transformation nodes in a chosen environment. This is perfect for verifying node configurations or checking the structure of your data pipelines during a migration. ## How It Works The bottom line is you get a natural language interface for your entire Coalesce data pipeline infrastructure. 1. Add the Coalesce integration to your AI client. 2. Input your API Token from the Coalesce Organization Settings. 3. Ask your agent to manage pipelines, check statuses, or trigger jobs. ## Frequently Asked Questions **Can I use the Coalesce MCP to trigger specific Snowflake jobs?** Yes, you can trigger specific jobs in any of your Coalesce environments using natural language. This allows you to kick off data transformations without needing to navigate the web dashboard. **How do I check the status of a pipeline using the Coalesce MCP?** You can simply ask your agent for the status of a specific run. It will check the progress and tell you if the pipeline is still running, succeeded, or failed. **Can the Coalesce MCP help me debug a failed data transformation?** Yes, it can pull detailed execution logs for failed jobs. This helps you identify specific errors in your Snowflake pipelines much faster than manual searching. **Can I see my different environments with the Coalesce MCP?** You can ask the agent to list all environments in your organization. It will provide a summary of the environments, their IDs, and the status of their most recent runs. **Is the Coalesce MCP compatible with my existing Snowflake setup?** Yes, because Coalesce is built for Snowflake, this Connector connects directly to your existing data transformation infrastructure to manage your pipelines. **How does the Coalesce MCP help with data engineering workflows?** It reduces the time spent on manual tasks like environment navigation and job triggering. Data engineers can stay in their workspace and manage their pipelines through simple chat commands. **How do I get my Coalesce API token?** In the Coalesce UI, go to **Organization Settings > API Tokens** and generate a new token. Copy it and paste it below. **Does this work with Snowflake?** Yes. Coalesce is built specifically for Snowflake. The API triggers jobs that run directly on your Snowflake instance. **What is a Node Selector?** A Node Selector lets you filter which transformation nodes to include in a run, based on name, type, or tags.