Skip to content

Troubleshooting

Preview

The AI Debug Assistant is currently in preview and may change in future releases.

Choosing the right path

If you're experiencing issues with the CFS Debug chat participant, we recommend switching to the CFS MCP debug server with any MCP-compatible AI client for maximum portability and future compatibility.

See Getting started for connection steps.

Port conflict

If the OS cannot assign a port, or you need a specific port:

  1. Go to Settings → Extensions → CodeFusion Studio → General, or search for (cfs.mcp.port).
  2. Change the Port (cfs.mcp.port) to a specific free port (for example, 3001 or 3002). Set it to 0 to let the OS assign a port automatically.
  3. Run (CFS) MCP: Stop Debug Server then (CFS) MCP: Start Debug Server.
  4. Update your AI client configuration to use the new port.

No active debug session

Most tools require an active debug session. Start debugging with F5 or via the Run menu before asking the assistant to inspect state or control execution. Alternatively, you can ask the assistant to start a debug session using the appropriate configuration name. For example: "Start debugging CFS: Debug with GDB and OpenOCD (ARM Embedded)".

No Copilot model available

If you see the error No Copilot model available. Please ensure GitHub Copilot is installed and authenticated, the @cfs-debug chat participant cannot access the required AI model. To resolve this:

  1. Ensure GitHub Copilot is installed and you are signed in
  2. Ensure your GitHub Copilot plan includes access to Claude models (such as Claude Sonnet 4.5 or Claude Opus 4.6)

CFS Debug Assistant tools not appearing in GitHub Copilot Chat

If the CFS Debug Assistant tools are not available in GitHub Copilot Chat:

  1. Check that your VS Code version is 1.96.0 or later (Help → About)
  2. Ensure GitHub Copilot is installed and you are signed in
  3. Confirm the CodeFusion Studio extension is installed and active

Claude Code cannot connect

If Claude Code cannot reach the MCP server:

  1. Run (CFS) MCP: Server Status in the Command Palette to confirm the MCP server is running
  2. Run (MCP) List Servers or claude mcp list to confirm cfs-debug is registered and check the URL and port are correct

  3. If the port has changed, re-register the server:

    claude mcp remove cfs-debug
claude mcp add --transport http cfs-debug http://localhost:<port>/mcp

Test the connection independently

Use the MCP Inspector to test the server directly:

npx @modelcontextprotocol/inspector

Select Streamable HTTP transport and Proxy connection type, then enter http://localhost:<port>/mcp (replace <port> with the port assigned at startup — check using (CFS) MCP: Server Status). You can browse all available tools, resources, and prompts, and execute them manually to verify everything is working before involving an AI client.

Additional information