Use these docs in this order:

  1. Get Studio running locally.
  2. Configure your MCP client.
  3. Wire your app to the SDK.
  4. Issue a pairing config and connect a device.
  5. Add tool suites only for the development workflows you actually need.
  6. Keep the whole setup out of public release builds.

Browse By Topic

Getting Started

First desktop-to-device setup

Install Studio, point your agent at the local MCP endpoint, add NuGet packages, and pair a device.

MCP Configuration

Configure your AI client

Connect Codex, Cursor, Claude Code, and IDE tooling to the local Ansight Studio MCP endpoint.

Studio

Desktop workflow

Register apps, issue configs, show pairing QR codes, inspect sessions, run analysis, and export captures.

SDK

Runtime integration and tool suites

See the currently available SDKs, then drill into the .NET package, pairing flows, configuration, and remote tools.

Security

Development-only best practices

Keep Ansight local, constrain tool exposure, manage pairing configs intentionally, and avoid shipping the SDK publicly.

Skills

Browse published skill URLs

Open the `/skills` tree, drill into `/skills/dotnet`, and copy the direct markdown URL you want to hand to an agent.

What Is Available Today

  • The detailed integration docs in this tree currently cover the .NET SDK.
  • The desktop Studio app hosts the local MCP bridge and pairing workflow.
  • Published prompt skills are indexed at /skills, with the current .NET subtree at /skills/dotnet.
  • Other SDKs are planned, but the .NET package is the only one documented here in depth right now.