Model Context Protocol

MiraBridge supports the Model Context Protocol (MCP) for extending the AI with custom tools and data sources.

What is MCP#

MCP is an open protocol that lets AI assistants connect to external tools and data sources through a standardized interface. MiraBridge implements MCP client support, allowing you to register MCP servers that provide additional capabilities.

Registering an MCP Server#

1. Open MiraBridge settings or the MCP toolbar button in the VS Code chat panel
2. Install a built-in server such as filesystem, git, memory, or context7
3. For custom servers, provide serverId, command, args, and env
4. The server's tools and resources become available to the AI

Auto-Discovery#

When enabled, MiraBridge can automatically discover MCP servers in your workspace based on configuration files.

Starter Configs#

Starter custom-server payloads live in examples/mcp/:

• examples/mcp/filesystem.json
• examples/mcp/git.json
• examples/mcp/memory.json
• examples/mcp/context7.json
• examples/mcp/github.json

These examples match the custom MCP server payload shape MiraBridge expects.

Use Cases#

• Database access — let the AI query your development database
• API integration — connect to internal APIs and services
• Custom tools — extend the AI with project-specific operations
• Documentation — provide the AI with access to internal docs

Next Steps#

• Workspace Rules — configure AI behavior
• Tools & Commands — built-in tools reference