move tag to align better with prefect version

update dependency versions in pyproject.toml

- Fixed flaky tests that were failing due to missing flow runs or wor…

…k queues

- Improved ID extraction logic with better regex patterns and JSON fallback parsing
- Tests now gracefully handle empty datasets instead of crashing

- Added utility function `extract_id_from_response()` for consistent ID parsing
- Better handling of UUID objects in API responses
- Tests work with existing data rather than requiring specific test fixtures

- Fixed `pytest.skip()` calls inside async contexts (can't skip mid-execution!)
- Better timeout handling for long-running operations
- More reliable test execution in CI/CD environments

- Added optional `name` parameter to `get_workspaces()` function
- Better documentation about filtering limitations across Prefect versions
- Maintains backward compatibility

- Removed unsupported `flow_name` parameter from flow run filtering tests
- Tests now use supported parameters only
- Fixed multiple tool requirement specification in test fixtures

- Changed from create-and-delete pattern to create-and-verify
- More reliable since deletion might not be immediately consistent
- Focuses on testing creation and retrieval rather than cleanup

- Enhanced ID extraction with multiple fallback strategies:
  1. UUID object pattern matching
  2. Simple string ID extraction
  3. JSON parsing with UUID object handling
- Better error handling and logging in test utilities

- Clearer test function names and descriptions
- Better comments explaining test strategies
- More informative logging during test execution

- Version bump to `0.2.2`
- Improved test comments and documentation
- Better handling of edge cases in test data parsing

---

*This release focuses on making the test suite more robust and reliable, especially when running against fresh Prefect instances with minimal data.*

- **BREAKING**: Replaced manual `get_all_functions()` pattern with `@…

…mcp.tool` decorators

- Tools are now auto-registered when modules are imported
- Cleaner, more maintainable code structure

- Updated FastMCP from `0.4.1` to `2.9.2`
- Added proper SSE (Server-Sent Events) support
- Server now binds to `0.0.0.0` for better accessibility

- Fixed `.dict()` deprecation warnings by switching to `.model_dump()`
- Updated all model serialization calls across the codebase

- Fixed deployment and flow filtering with proper filter objects
- No more `unexpected keyword argument` errors
- Filtering actually works as expected now

- Fixed variable deletion by using correct API method (`delete_variable_by_name()`)
- Improved variable creation with proper model validation

- **Prefect**: `3.2.15.dev8` → `3.4.6`
- **Docker**: Updated to `prefect:3.3.3-python3.12`
- Various Python package updates for security and compatibility
- Added `pytest-asyncio` for better async testing

- Added try-catch blocks with meaningful error messages
- Better error responses instead of silent failures

- Added comprehensive test suite with pytest
- Proper async test configuration
- Test scripts with examples

- Added usage examples in README
- Better configuration documentation
- Improved setup instructions

1. **Tool Registration**: Must use `@mcp.tool` decorators instead of `get_all_functions()`
2. **Model Methods**: Replace `.dict()` with `.model_dump()` calls
3. **Variable API**: Use new deletion method name
4. **Dependencies**: Several version bumps may require updates

- `/health` endpoint for monitoring
- Better SSE transport support
- Improved configuration handling
- More robust error responses

- Update tool definitions to use decorators
- Check any custom Pydantic model usage
- Update dependency versions
- Test thoroughly after upgrade

---

*This is a significant architectural update that modernizes the codebase and fixes several long-standing issues. ( from like 90 days ago, like so last year or something )*