# Northflank MCP

> Northflank MCP lets you manage your entire developer cloud infrastructure directly through conversation. Orchestrate microservices, deploy new code builds, audit complex background jobs, and provision or delete entire project ecosystems without touching a dashboard or running a single CLI command.

## Overview
- **Category:** ship-it
- **Price:** Free
- **Tags:** microservices, cloud-deployment, ci-cd, infrastructure-orchestration, backend-hosting

## Description

Control your enterprise development platform using natural language. This connection gives your agent full oversight of your microservice orchestration and continuous deployment pipelines across multiple cloud regions. You can list every managed project and get detailed metadata, checking things like which geographic data centers (AWS, GCP, Azure) are connected. Need to restart a service? Simply ask the agent to cycle container replicas for specific applications. The MCP also handles operational tasks like listing isolated batch jobs or auditing secret groups—verifying environment variables across different virtual private cloud boundaries. When you connect this via Vinkius, your AI client treats Northflank as just another tool in its belt, letting you manage complex infrastructure from anywhere.

## Tools

### create_project
Creates a completely new, isolated Northflank Project space for development work.

### delete_project
Permanently removes an entire project boundary and all associated microservices and running jobs.

### get_project
Retrieves detailed regional metadata about a specific Northflank Project, including its cloud boundaries.

### get_service
Fetches the exact structural anatomy and scaling needs of a single running application service.

### list_jobs
Lists all isolated batch and cron jobs that run within a specific project.

### list_projects
Fetches the mandatory list of top-level organizational Project IDs.

### list_secrets
Lists all secret group dictionaries, showing environment variables mapped into running services or jobs.

### list_services
Retrieves a list of explicit application or microservice instances inside a project.

### restart_service
Gracefully cycles the container replicas for a specific service to clear transient memory buildup.

### trigger_build
Sends a command to Northflank builders to compile and deploy the latest source code.

## Prompt Examples

**Prompt:** 
```
List all services in the 'api-gateway' project
```

**Response:** 
```
I've retrieved the services for 'api-gateway' (ID: gate-123). Highlights include 'Auth-Service' (Status: Running), 'Payment-Processor' (Status: Running), and 'Notification-Node'. Would you like to check the resource allocation for the auth service?
```

**Prompt:** 
```
Trigger a fresh build for service 'gate-123' in project 'proj-abc'
```

**Response:** 
```
Triggering production build… Success. I've signaled Northflank to pull the latest code from GitHub and start the build cycle for 'Auth-Service'. I'll monitor the progress and let you know once the new image is live.
```

**Prompt:** 
```
Show me the secret groups configured in my project
```

**Response:** 
```
Retrieving secret vaults… I've identified 2 logical groups: 'Production-Secrets' (Group ID: sec-987) and 'Shared-API-Keys'. I can provide the names of the environment variables mapped within these groups if you'd like.
```

## Capabilities

### Manage Project Lifecycle
Provision new isolated project spaces or permanently tear down existing microservice ecosystems.

### Audit Service Health
Retrieve the precise resource allocation and structural details for any running application service.

### Control Deployments
Manually trigger fresh continuous integration builds to update production assets or verify recent code merges.

### Restart Services
Gracefully cycle container replicas for a specific service, clearing accumulated transient memory and restoring normal performance timing.

### Monitor Background Processes
List and inspect isolated batch or cron jobs to track periodic tasks like heavy database aggregations.

### Verify Credentials
Access metadata for logical secret groups, confirming environment variable mappings across your cloud boundaries.

## Use Cases

### Investigating Production Outages
A payment service is running slow. Instead of logging into three different consoles, the engineer asks their agent to `get_service` for that specific application. The agent returns immediate data on CPU throttling and RAM allocation boundaries, pointing directly to the resource bottleneck.

### Preparing a New Team Environment
A new team needs a sandbox environment. Instead of manually setting up networking rules, the platform lead asks the agent to `create_project`. The system provisions an isolated project space instantly, ready for development.

### Auditing Security Changes
Before migrating data, the security team needs to confirm where database connection strings are used. They ask the agent to use `list_secrets` and get a map of all environment variable types across different operational zones.

### Verifying Scheduled Tasks
A nightly report is failing, but nobody knows which cron job handles it. The developer asks the agent to `list_jobs`, retrieving a list of isolated batch processes and confirming the exact scheduled task that needs fixing.

## Benefits

- Instant Service Audit: Instead of manually checking dashboards, ask your agent to use `get_service` to instantly verify a service’s structural anatomy and current resource boundaries.
- Rapid Deployments: When code changes, you don't navigate to the CI/CD dashboard. Just prompt your agent to `trigger_build`, starting the deployment immediately.
- Clean Up Effortlessly: Need to decommission an old testing environment? Use `delete_project` to permanently tear down a project and all its microservices in one command.
- Proactive Maintenance: Don't wait for failure. Tell your agent to use `restart_service` on critical applications to cycle replicas and clear memory buildup proactively.
- Security Visibility: Quickly verify credentials by asking the agent to use `list_secrets`, checking environment variable mappings across all connected VPC boundaries.
- Full Oversight: Need a quick inventory? Use `list_projects` or `list_services` to get an immediate, comprehensive list of every resource in your account.

## How It Works

The bottom line is you can use natural conversation to execute complex DevOps operations that used to require multiple specialized dashboards and terminal commands.

1. Subscribe to this MCP on Vinkius and provide the required Northflank API Token.
2. Connect your preferred AI client (like Cursor or Claude) to the catalog, granting it access to the tool.
3. Ask your agent a natural language question—for example, 'List all active microservices in Project X'—and let it perform the necessary cloud actions.

## Frequently Asked Questions

**How do I list all services using Northflank (Developer Cloud & Orchestration) MCP?**
You use the `list_services` tool. This fetches a complete roster of every explicit application or microservice instance running inside your current project.

**What is the difference between `get_project` and `list_projects` in Northflank (Developer Cloud & Orchestration) MCP?**
`list_projects` gives you a list of all top-level organizational Project IDs. `get_project` allows you to dive deep into one specific project ID to retrieve its detailed regional metadata.

**Can I restart my service using Northflank (Developer Cloud & Orchestration) MCP?**
Yes, you use the `restart_service` tool. This command gracefully cycles container replicas for a specific application, clearing out any transient memory buildup.

**I need to find all my secrets, which tool should I use in Northflank (Developer Cloud & Orchestration) MCP?**
Use `list_secrets`. This lists all secret group dictionaries and shows you exactly which environment variables are mapped into your running services or jobs.

**What if I want to tear down an entire project? Which tool handles that in Northflank (Developer Cloud & Orchestration) MCP?**
Use the `delete_project` tool. This permanently removes a Project and all its cascading microservices, effectively cleaning up the entire ecosystem.