# Bunny.net MCP

> Bunny.net MCP connects your AI agent directly to Bunny.net edge infrastructure. Use natural language commands to manage global CDN Pull Zones, purge caching rules instantly, and map cloud storage volumes without touching a CLI or portal. It lets you provision new acceleration links, inspect deep file structures, and delete entire zones—all from conversation.

## Overview
- **Category:** developer-tools
- **Price:** Free
- **Tags:** cdn, edge-computing, cloud-storage, caching, network-acceleration

## Description

Managing your content delivery network shouldn't require logging into three different dashboards and running messy shell scripts. This MCP connects your AI agent to Bunny.net, letting you control every aspect of your edge infrastructure using only natural conversation. You can create new global reverse proxy links or delete old ones with a simple prompt. Need to force users onto fresh code? Just ask the agent to purge cache on a specific zone. The whole thing—from provisioning storage volumes to inspecting binary assets deep inside—is accessible right through the Vinkius catalog, making your AI client your core edge configuration wizard.

## Tools

### add_pull_zone
Creates a new, fast CDN reverse proxy domain for global content distribution.

### add_storage_zone
Sets up an empty replication volume in the cloud storage area.

### delete_pull_zone
Instantly removes an active CDN acceleration mapping, dropping global traffic connections.

### delete_storage_zone
Completely wipes and deletes an entire existing CDN storage zone.

### get_pull_zone
Retrieves detailed metadata about a specific global content delivery proxy link.

### get_storage_zone
Inspects the credentials and origin points for a precise storage zone.

### list_files
Lists all uploaded binary assets inside a specified storage zone, including their byte size.

### list_pull_zones
Identifies and lists all active CDN accelerated delivery domains (Pull Zones).

### list_storage_zones
Finds and reports on all physical Edge Storage origin folders.

### purge_pull_zone
Forces an immediate cache invalidation, making the CDN re-fetch content directly from the source.

## Prompt Examples

**Prompt:** 
```
List all my current Pull Zones and their original targets.
```

**Response:** 
```
You have 2 Pull Zones explicitly active. ID 12040 runs 'prod-app-cdn' targeting your AWS S3 bucket. ID 19302 runs 'staging-cdn' pointing to your beta server IP. Do you need to examine their advanced security configurations?
```

**Prompt:** 
```
My site design looks old, the JS didn't update. Purge Pull Zone ID 12040 right now.
```

**Response:** 
```
I've sent the global invalidation command. The CDN cache on Pull Zone 12040 has been instantly purged. Your edge nodes are now re-fetching the JS origin bundles actively. You're good to go.
```

**Prompt:** 
```
What files are currently inside the '/assets/' path of our 'main-storage-x' Edge Storage zone?
```

**Response:** 
```
I checked your NVMe edge cluster. Within '/assets/', I found 'hero_bg.webp' (1.2 MB) and 'compiled_main.css' (900 KB). Shall I provide the cryptographic checksums?
```

## Capabilities

### Manage CDN Links
Create new global reverse proxy domains or delete existing ones for instant traffic control.

### Control Caching Rules
Force immediate cache invalidations across specific zones, guaranteeing users hit the latest content.

### Inspect Storage Data
List files and check metadata inside your raw origin storage volumes via FTP parameters.

### Provision & Wipe Zones
Bootstrap new empty replication volumes or vaporize existing CDN storage zones entirely.

## Use Cases

### The Stale JS Bundle Problem
A front-end dev pushes a new version of their Javascript bundle but the live site shows old code. They prompt the agent: 'Purge Pull Zone ID 12040 now.' The MCP runs purge_pull_zone, instantly forcing all global edge nodes to re-fetch the fresh script.

### The New Regional Endpoint
A team needs a new CDN endpoint for their beta region. Instead of manual setup, they simply ask the agent to execute add_pull_zone, provisioning a radically fast reverse proxy domain instantly.

### Archiving an Old Data Set
An old project's storage volume is no longer needed. The ops engineer uses delete_storage_zone on the specific zone ID to vaporize it completely, freeing up the namespace and cleaning the registry.

### Validating Content Integrity
A media owner suspects a file was corrupted during upload. They use list_files on the storage zone and check the cryptographic checksums against their manifest to prove data integrity.

## Benefits

- Force fresh content delivery: Use purge_pull_zone to instantly invalidate the cache on any zone, bypassing slow Time-To-Live rules for immediate updates.
- Full infrastructure control: Easily add_pull_zone or add_storage_zone when you need a new endpoint without manually configuring network portals.
- Deep inspection capability: Use list_files to enumerate uploaded binary assets and check their cryptographic hash footprints inside the storage zone.
- Rapid cleanup: When a proxy link is obsolete, delete_pull_zone vaporizes it completely, stopping all global traffic flowing through that mapping.
- Visibility into resources: Run list_storage_zones or list_pull_zones to get an immediate inventory of every active and potential data boundary.

## How It Works

The bottom line is you talk to your AI client; it talks to Bunny.net and makes the changes.

1. Subscribe to this MCP and enter your master Bunny API Key along with the necessary zone tokens.
2. Connect your preferred AI client (Claude, Cursor, etc.) through Vinkius.
3. Issue a direct command in natural language—your agent handles the complex interaction with the CDN edge structure.

## Frequently Asked Questions

**How do I list all my current CDN-accelerated domains using the Bunny.net MCP?**
Use list_pull_zones to see a full inventory of your global delivery domains, including their hostnames and targets.

**What is the difference between add_storage_zone and get_storage_zone?**
add_storage_zone boots up an empty replication volume. Conversely, get_storage_zone lets you inspect existing volumes to check their origin PoPs and credentials.

**Can I delete a zone using the purge_pull_zone tool?**
No. purge_pull_zone only invalidates the cache. To remove the actual connection or mapping, you must use delete_pull_zone.

**How do I find out what files are inside my Edge Storage zone?**
You run list_files. This tool requires a specific zone password but returns explicit byte lengths and hash footprints for all stored assets.

**What kind of credentials or setup details do I need before running a command like `add_pull_zone`?**
You must provide your master Bunny API Key and standard zone tokens. These keys authorize the MCP to make changes to your CDN infrastructure. The system validates these against your account permissions before executing any write operation.

**Using `get_pull_zone`, what specific advanced metadata can I extract about an existing domain?**
You retrieve detailed configuration like enabled SSL settings, bandwidth limits, and any custom Edge Rules intercepting incoming requests. This lets you audit the structural integrity of a specific Pull Zone.

**If my Storage Zone has files, does `delete_storage_zone` require me to empty it first?**
No, the tool is designed to handle this directly. It can force recursive wipes, freeing the entire namespace back into the registry. Use caution, as this action vaporizes the zone entirely.

**When should I use `add_pull_zone` instead of modifying an existing one?**
You run `add_pull_zone` when you need to provision a completely new, distinct CDN reverse proxy domain. This establishes a separate delivery channel for media that isn't currently mapped anywhere.

**Can my AI agent instantly purge a CDN Pull Zone after a release?**
Yes. Instead of logging into the admin portal manually, just tell your agent you deployed the frontend. It will grab the ID of your main Pull Zone and blast the purge_pull_zone command immediately.

**Does it track new files uploaded directly to my Storage Zone?**
Yes! Provide the API password and string path, and the agent lists precise object files inside the folder natively. E-commerce sites use this to verify if static images successfully pushed to the edge before updating database paths.

**Can it provision new Pull Zones from scratch?**
Absolutely. You can request a fresh caching domain, pass the target origin URL (like an S3 bucket), and the AI builds your CDN layer generating the specific `*.b-cdn.net` URL response for you.