flutterdocs_mcp 0.1.1

Flutter/Dart API documentation for AI assistants and tools.

example/example.md

Using flutterdocs_mcp

Below are examples and tips for configuring flutterdocs_mcp with various hosts and using it effectively.

VS Code

To use flutterdocs_mcp with VS Code, add the following to your mcp.json:

{
	"servers": {
		"flutterdocs": {
			"type": "stdio",
			"command": "flutterdocs_mcp",
			"args": ["--db", "/path/to/flutterdocs.db"]
		}
	}
}

The MCP server name is flutterdocs as configured above. This is the name that will appear in VS Code tools and that should be used in chat sessions.

Tips:

• Thoroughly review the VS Code documentation on configuring, managing, and using MCP servers.
• Select Configure Tools in chat sessions to confirm the MCP server name is selected.
• Reference the MCP server name in chat sessions to ensure the agent uses it to formulate responses.
  • summarize materialpageroute usage from flutterdocs
    • explain flutter slivers. reference flutterdocs to verify api details.
• Attach resources to chat sessions by selecting Add Context > MCP Resources or using the MCP: Browse Resources command and then choosing one of the resource templates associated with the MCP server name.
• My experience is that GPT-5 mini struggles with MCP servers. Some others may as well; I did not test all available models.

MCP Inspector

MCP Inspector is a tool for testing and debugging MCP servers. To use flutterdocs_mcp with MCP Inspector first install Node.js and npm on your system and then run:

npx @modelcontextprotocol/inspector flutterdocs_mcp -- --db /path/to/flutterdocs.db