* Catch more openapi changes
* Simplify middleware list operations to use lists directly
- Updated middleware signatures to work with list[Tool], list[Resource], etc instead of wrapper objects
- Removed unnecessary ListToolsResult, ListResourcesResult wrapper types
- Updated documentation to show simpler list-based filtering pattern
- Added tests for list-based middleware filtering
Closes#1121🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
---------
Co-authored-by: Claude <noreply@anthropic.com>
* Fix OpenAPI deepObject style parameter encoding
Add support for deepObject style with explode=true to properly serialize
object parameters using bracket notation (param[key]=value) instead of
JSON strings. Fixes#1114.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Create README_OPENAPI.md
---------
Co-authored-by: Claude <noreply@anthropic.com>
Resolves parameter conflicts when same name appears in path/query/header
and request body by adding double underscore suffixes (id__path, id__query).
Body parameters keep original names for natural REST API patterns.
Closes#1100🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Prevents enable/disable operations on mirrored components (tools, resources, prompts retrieved from proxy servers). Users must create local copies with .copy() and add them to their server to modify them. Local components take precedence over mirrored ones.
Closes#1102🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Resolves issue where tools modified by mcp_component_fn were registered
with original names but accessible with modified names, causing "Unknown tool"
errors. Now tools are registered using their final modified names.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-authored-by: Claude <noreply@anthropic.com>
* Refactor Client session state with ClientSessionState dataclass
Fixes#1068 by introducing ClientSessionState to encapsulate session management
attributes, simplifying Client.new() and preventing concurrent proxy client
context mixing through client factory pattern.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Update proxy documentation for new client factory pattern
Documents the new session management behavior, client_factory parameter,
and concurrent operation safety introduced in v2.10.3 to fix#1068.
* Remove deprecation of client parameter and update documentation
- Undeprecated FastMCPProxy client parameter
- Made client_factory the advanced option for custom control
- Added detailed explanation of how client factories work internally
- Removed outdated note about proxy feature limitations
- Removed fabricated Advanced Usage section
* Re-do proxy documentation
* Fix deprecated client parameter to provide session isolation
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
---------
Co-authored-by: Claude <noreply@anthropic.com>
* Add output schema support for OpenAPI tools
Implement automatic output schema extraction from OpenAPI responses for
FastMCP tools, addressing issue #1070. Tools generated from OpenAPI specs
now have meaningful output schemas instead of null, improving agent
efficiency when working with structured API responses.
- Add extract_output_schema_from_responses() function to parse response schemas
- Enhance OpenAPITool to accept and use output schemas with proper wrapping
- Automatically wrap non-object responses to comply with MCP requirements
- Include schema definitions and compress unused ones
- Add comprehensive test suite covering object, array, and primitive responses
- Update existing tests to handle structured output with proper typing
- Maintain backward compatibility for specs without response schemas
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Move output schema tests to dedicated test file
Reorganize output schema tests into test_openapi_output_schemas.py to keep
the main test_openapi.py file focused and smaller, as requested. Added two
additional tests for schema definitions handling.
- Move all extract_output_schema_from_responses() tests to new file
- Add tests for schema definitions inclusion
- Remove output schema imports from main test file
---------
Co-authored-by: Claude <noreply@anthropic.com>
Single-element lists like [1] were being incorrectly unwrapped to "1"
in unstructured content while multi-element lists remained as lists.
This created inconsistent behavior where the structure was lost for
single items.
This fix ensures lists always preserve their structure in unstructured
content regardless of length, making behavior consistent and predictable.
Also removes pretty-printing from JSON serialization for more compact
output across tools, prompts, and resources.
Fixes#1064🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Adds test coverage for specific differences between OpenAPI 3.0 and 3.1 that can cause validation issues, including the scenario reported in GitHub issue #1021.
Test coverage includes:
- OpenAPI 3.0 boolean exclusiveMaximum format vs 3.1 numeric format
- OpenAPI 3.0 nullable vs 3.1 type array format
- Complex schemas with $defs and multiple exclusive constraints
- Edge cases with both exclusiveMaximum and exclusiveMinimum
These tests serve as regression prevention and document expected behavior for both OpenAPI versions.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Because the server now validates inputs, its no longer viable to provide a JSON string where e.g. an array is expected, so we can entirely remove FastMCP's "legacy json parsing" support from 1.x.
This is a breaking change (for legacy opt-in behavior)
Fixes breaking changes from mcp-python-sdk PR #982 which updated
BearerAuthBackend to use TokenVerifier protocol instead of OAuth providers.
Changes:
- Add verify_token() method to BearerAuthProvider implementing TokenVerifier protocol
- Add verify_token() method to InMemoryOAuthProvider implementing TokenVerifier protocol
- Update BearerAuthBackend usage in setup_auth_middleware_and_routes() to pass TokenVerifier
- Add comprehensive unit tests for TokenVerifier implementations
- Add integration tests for BearerAuthBackend with TokenVerifier
- Add tests for HTTP auth setup functions
All existing functionality preserved with full backwards compatibility.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Replace hasattr(param_in, 'value') with isinstance(param_in, Enum)
for more robust enum detection as suggested in review.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Fixes issue where ParameterLocation enum values from openapi_pydantic
were not properly converted to strings, causing validation errors.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Implements timing, logging, rate limiting, and error handling middleware to showcase
FastMCP's middleware capabilities as a headline feature. Each middleware includes:
- Production-ready implementations with full configurability
- Comprehensive unit and integration tests with real FastMCP servers
- Updated documentation with teaching examples and production usage patterns
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Add proper null checks and type assertions for prompt argument
handling in tests to satisfy pyright strict typing.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Change from 'Arguments must be strings conforming to this JSON schema'
to 'Provide as a JSON string matching the following schema' for clearer
instruction to LLMs about string format requirements.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Fix ValueError -> PromptError for consistent error handling
- Add automatic JSON schema descriptions to non-string prompt arguments
- Include comprehensive tests for argument description enhancement
- Verify enhanced descriptions are visible via MCP protocol
This helps developers understand the expected string format for complex
types when calling prompts from MCP clients.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Exposes MCP session ID from HTTP headers to enable Redis-based data sharing
between tools. Returns None for stdio/memory transports as expected.
Fixes#875, addresses python-sdk#986, python-sdk#942, python-sdk#485
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
