Skip to main content
Glama

FastAPI-MCP

CHANGELOG.md3.99 kB
# Changelog All notable changes to this project will be documented in this file. The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). ## [0.2.0] ### Changed - Complete refactor from function-based API to a new class-based API with `FastApiMCP` - Explicit separation between MCP instance creation and mounting with `mcp = FastApiMCP(app)` followed by `mcp.mount()` - FastAPI-native approach for transport providing more flexible routing options - Updated minimum MCP dependency to v1.6.0 ### Added - Support for deploying MCP servers separately from API service - Support for "refreshing" with `setup_server()` when dynamically adding FastAPI routes. Fixes [Issue #19](https://github.com/tadata-org/fastapi_mcp/issues/19) - Endpoint filtering capabilities through new parameters: - `include_operations`: Expose only specific operations by their operation IDs - `exclude_operations`: Expose all operations except those with specified operation IDs - `include_tags`: Expose only operations with specific tags - `exclude_tags`: Expose all operations except those with specific tags ### Fixed - FastAPI-native approach for transport. Fixes [Issue #28](https://github.com/tadata-org/fastapi_mcp/issues/28) - Numerous bugs in OpenAPI schema to tool conversion, addressing [Issue #40](https://github.com/tadata-org/fastapi_mcp/issues/40) and [Issue #45](https://github.com/tadata-org/fastapi_mcp/issues/45) ### Removed - Function-based API (`add_mcp_server`, `create_mcp_server`, etc.) - Custom tool support via `@mcp.tool()` decorator ## [0.1.8] ### Fixed - Remove unneeded dependency. ## [0.1.7] ### Fixed - [Issue #34](https://github.com/tadata-org/fastapi_mcp/issues/34): Fix syntax error. ## [0.1.6] ### Fixed - [Issue #23](https://github.com/tadata-org/fastapi_mcp/issues/23): Hide handle_mcp_connection tool. ## [0.1.5] ### Fixed - [Issue #25](https://github.com/tadata-org/fastapi_mcp/issues/25): Dynamically creating tools function so tools are useable. ## [0.1.4] ### Fixed - [Issue #8](https://github.com/tadata-org/fastapi_mcp/issues/8): Converted tools unuseable due to wrong passing of arguments. ## [0.1.3] ### Fixed - Dependency resolution issue with `mcp` package and `pydantic-settings` ## [0.1.2] ### Changed - Complete refactor: transformed from a code generator to a direct integration library - Replaced the CLI-based approach with a direct API for adding MCP servers to FastAPI applications - Integrated MCP servers now mount directly to FastAPI apps at runtime instead of generating separate code - Simplified the API with a single `add_mcp_server` function for quick integration - Removed code generation entirely in favor of runtime integration ### Added - Main `add_mcp_server` function for simple MCP server integration - Support for adding custom MCP tools alongside API-derived tools - Improved test suite - Manage with uv ### Removed - CLI interface and all associated commands (generate, run, install, etc.) - Code generation functionality ## [0.1.1] - 2024-07-03 ### Fixed - Added support for PEP 604 union type syntax (e.g., `str | None`) in FastAPI endpoints - Improved type handling in model field generation for newer Python versions (3.10+) - Fixed compatibility issues with modern type annotations in path parameters, query parameters, and Pydantic models ## [0.1.0] - 2024-03-08 ### Added - Initial release of FastAPI-MCP - Core functionality for converting FastAPI applications to MCP servers - CLI tool for generating, running, and installing MCP servers - Automatic discovery of FastAPI endpoints - Type-safe conversion from FastAPI endpoints to MCP tools - Documentation preservation from FastAPI to MCP - Claude integration for easy installation and use - API integration that automatically makes HTTP requests to FastAPI endpoints - Examples directory with sample FastAPI application - Basic test suite

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/nguyendinhsinh361/fastapi-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server