

The new Avalonia DevTools MCP server is available today in 2.2.0-beta2. It's a programmatic bridge between AI agents and your running Avalonia applications, essentially DevTools, but controllable by AI.
The Problem
Large-scale app porting work can be tedious. Whether you're migrating thousands of views, recreating designs from mockups, or fine-tuning layouts to match specifications, the work isn't technically complex; it's just time-consuming and repetitive. The kind of work that burns out good engineers because they're spending their days tweaking margins and alignment instead of solving interesting problems.
The Solution: Programmatic DevTools
The Avalonia DevTools MCP server (`avalonia_devtools`) gives AI agents direct access to inspect and manipulate your running Avalonia applications.

Here's what it provides:
Connection & Discovery
Connect to any running Avalonia app
List and switch between multiple running instances
Tree Inspection
Traverse visual, logical, or merged element trees
Search for elements by type or x:Name
Walk ancestor chains from any node to root
Property & Style Manipulation
Read property values from any UI element
Set properties at runtime with full XAML value support
Inspect applied styles and setters
Query and toggle pseudo-classes (`
:pointerover`, `:pressed`, `:focus`, etc.)
Resources & Assets
Access resources scoped to the app or specific nodes
List embedded assets
Download assets via `avares://` URLs
UI Interaction
Perform actions: Focus, Enable, Disable, BringIntoView
Capture PNG screenshots of any UI element
The Workflow
Here's a typical workflow:
Feed an AI agent a screenshot or design mockup
The agent generates initial Avalonia XAML
Using the MCP server, it connects to the running app
It searches the visual tree to find the elements it created
It inspects property values and compares screenshots
It iterates, adjusting margins, colours, layouts, until it matches
Move to the next view
What would take 30-60 minutes of manual tweaking now takes minutes, with the AI handling the tedious property adjustments while you focus on architecture and logic.
Real-World Example
Last week, I was in Montreal demonstrating this with Devolutions, which is porting its Remote Desktop Manager to Avalonia. They have approximately 5,000 views to migrate, which is a huge amount of repetitive UI work!
We took one of their existing views, gave Claude a screenshot, and let it work. Within a few minutes, a pixel-perfect recreation, running in Avalonia, was ready for testing.

The value isn't just speed; it's the quality of the iteration loop. The agent isn't guessing. It's inspecting, measuring, comparing. It can see issues and fix them proactively. It doesn’t require you to reprompt because it didn’t get it right. It can verify that everything is looking correct before moving on.
This changes the economics of large-scale app porting work!
Available Today
The Avalonia MCP server ships with DevTools 2.2.0-beta2, available now. If you're an Accelerate customer running the latest beta, you can start using it immediately.
The new Avalonia DevTools MCP server is available today in 2.2.0-beta2. It's a programmatic bridge between AI agents and your running Avalonia applications, essentially DevTools, but controllable by AI.
The Problem
Large-scale app porting work can be tedious. Whether you're migrating thousands of views, recreating designs from mockups, or fine-tuning layouts to match specifications, the work isn't technically complex; it's just time-consuming and repetitive. The kind of work that burns out good engineers because they're spending their days tweaking margins and alignment instead of solving interesting problems.
The Solution: Programmatic DevTools
The Avalonia DevTools MCP server (`avalonia_devtools`) gives AI agents direct access to inspect and manipulate your running Avalonia applications.

Here's what it provides:
Connection & Discovery
Connect to any running Avalonia app
List and switch between multiple running instances
Tree Inspection
Traverse visual, logical, or merged element trees
Search for elements by type or x:Name
Walk ancestor chains from any node to root
Property & Style Manipulation
Read property values from any UI element
Set properties at runtime with full XAML value support
Inspect applied styles and setters
Query and toggle pseudo-classes (`
:pointerover`, `:pressed`, `:focus`, etc.)
Resources & Assets
Access resources scoped to the app or specific nodes
List embedded assets
Download assets via `avares://` URLs
UI Interaction
Perform actions: Focus, Enable, Disable, BringIntoView
Capture PNG screenshots of any UI element
The Workflow
Here's a typical workflow:
Feed an AI agent a screenshot or design mockup
The agent generates initial Avalonia XAML
Using the MCP server, it connects to the running app
It searches the visual tree to find the elements it created
It inspects property values and compares screenshots
It iterates, adjusting margins, colours, layouts, until it matches
Move to the next view
What would take 30-60 minutes of manual tweaking now takes minutes, with the AI handling the tedious property adjustments while you focus on architecture and logic.
Real-World Example
Last week, I was in Montreal demonstrating this with Devolutions, which is porting its Remote Desktop Manager to Avalonia. They have approximately 5,000 views to migrate, which is a huge amount of repetitive UI work!
We took one of their existing views, gave Claude a screenshot, and let it work. Within a few minutes, a pixel-perfect recreation, running in Avalonia, was ready for testing.

The value isn't just speed; it's the quality of the iteration loop. The agent isn't guessing. It's inspecting, measuring, comparing. It can see issues and fix them proactively. It doesn’t require you to reprompt because it didn’t get it right. It can verify that everything is looking correct before moving on.
This changes the economics of large-scale app porting work!
Available Today
The Avalonia MCP server ships with DevTools 2.2.0-beta2, available now. If you're an Accelerate customer running the latest beta, you can start using it immediately.
The new Avalonia DevTools MCP server is available today in 2.2.0-beta2. It's a programmatic bridge between AI agents and your running Avalonia applications, essentially DevTools, but controllable by AI.
The Problem
Large-scale app porting work can be tedious. Whether you're migrating thousands of views, recreating designs from mockups, or fine-tuning layouts to match specifications, the work isn't technically complex; it's just time-consuming and repetitive. The kind of work that burns out good engineers because they're spending their days tweaking margins and alignment instead of solving interesting problems.
The Solution: Programmatic DevTools
The Avalonia DevTools MCP server (`avalonia_devtools`) gives AI agents direct access to inspect and manipulate your running Avalonia applications.

Here's what it provides:
Connection & Discovery
Connect to any running Avalonia app
List and switch between multiple running instances
Tree Inspection
Traverse visual, logical, or merged element trees
Search for elements by type or x:Name
Walk ancestor chains from any node to root
Property & Style Manipulation
Read property values from any UI element
Set properties at runtime with full XAML value support
Inspect applied styles and setters
Query and toggle pseudo-classes (`
:pointerover`, `:pressed`, `:focus`, etc.)
Resources & Assets
Access resources scoped to the app or specific nodes
List embedded assets
Download assets via `avares://` URLs
UI Interaction
Perform actions: Focus, Enable, Disable, BringIntoView
Capture PNG screenshots of any UI element
The Workflow
Here's a typical workflow:
Feed an AI agent a screenshot or design mockup
The agent generates initial Avalonia XAML
Using the MCP server, it connects to the running app
It searches the visual tree to find the elements it created
It inspects property values and compares screenshots
It iterates, adjusting margins, colours, layouts, until it matches
Move to the next view
What would take 30-60 minutes of manual tweaking now takes minutes, with the AI handling the tedious property adjustments while you focus on architecture and logic.
Real-World Example
Last week, I was in Montreal demonstrating this with Devolutions, which is porting its Remote Desktop Manager to Avalonia. They have approximately 5,000 views to migrate, which is a huge amount of repetitive UI work!
We took one of their existing views, gave Claude a screenshot, and let it work. Within a few minutes, a pixel-perfect recreation, running in Avalonia, was ready for testing.

The value isn't just speed; it's the quality of the iteration loop. The agent isn't guessing. It's inspecting, measuring, comparing. It can see issues and fix them proactively. It doesn’t require you to reprompt because it didn’t get it right. It can verify that everything is looking correct before moving on.
This changes the economics of large-scale app porting work!
Available Today
The Avalonia MCP server ships with DevTools 2.2.0-beta2, available now. If you're an Accelerate customer running the latest beta, you can start using it immediately.






