Fixing tests/linting and image generation #1

rlorenzo · 2025-08-04T01:49:55Z

I was looking for an MCP server to use in my project to generate art assets and found your project. I couldn't get it to work out of the box, so here's a patch that got it working for me.

Some highlights:

Update environment variables to provider-namespaced format (PROVIDERS____)
Switch Gemini provider from Generative AI API to Vertex AI (to support Imagen-4)
Fix gpt-image-1 model capabilities (remove unsupported style parameter)
Add Claude Code MCP integration with start-mcp.sh script
Add .gitignore file with Python, IDE, and build artifacts exclusions
Fixed all failing tests and linting issues

- Add .gitignore file with Python, IDE, and build artifacts exclusions - Improve code formatting and linting compliance across all modules - Fixed failing tests - Removing API key logging

- Update environment variables to provider-namespaced format (PROVIDERS__*__*) - Switch Gemini provider from Generative AI API to Vertex AI with service account auth - Fix gpt-image-1 model capabilities (remove unsupported style parameter) - Add Claude Code MCP integration with start-mcp.sh script - Update Docker configuration and documentation with Vertex AI setup instructions 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>

…prove flexibility **Fix invalid test data creation** - Replace corrupted image data (bytes * 1000) with proper PIL-generated PNG images - Add create_larger_test_image() helper with comprehensive test coverage **Enhance image format detection** - Add _detect_image_format() function using byte signature analysis - Replace hardcoded PNG assumption with actual format detection **Optimize storage cleanup performance** - Replace O(n²) directory scanning with O(n) incremental size tracking - Calculate total size once, then decrement as files are deleted **Make provider validation more flexible** - Only validate providers that are actually enabled (not both required) - ImageGenerationTool works with any enabled provider combination - Update tests for new flexible validation behavior All 104 tests passing with improved performance and correctness. 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>

- Add defensive hasattr checks in ImageEditingTool constructor with detailed error message - Extract PNG signature to module constant in test_storage.py - Extract image format magic numbers to named constants in validators.py - Improve storage directory error messages with specific paths and solutions 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>

- Improve comment clarity about MagicMock vs AsyncMock usage in test_tools.py - Move inline import statements to top of test methods in test_storage.py - Improve WebP signature detection readability with extracted is_webp variable - Add comprehensive test suite for image format detection with magic numbers - Verify environment variable names are already correct in image_generation.py 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>

…feedback - Remove unnecessary parentheses in test assertion for cleaner code style - Clarify size limit comment with explicit GB to MB conversion in test_storage.py - Use double quotes consistently throughout conftest.py for string literals - Extract WebP format detection to separate _is_webp_format() helper function - Add comprehensive tests for WebP format detection helper function - Update store_image() docstring to mark as deprecated with clearer guidance - Extract OpenAI settings validation to _validate_openai_settings() helper method - Add dedicated test for OpenAI settings validation helper method 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>

- Simplify WebP return statement to single line for better readability - Enhance deprecation notice with timeline (v1.5.0, Q4 2024) and migration examples - Extract OpenAI validation to robust static method with comprehensive error handling - Add extensive test coverage for static validation method with edge cases - Fix critical package name mismatch in pyproject.toml (image-gen-mcp → gpt_image_mcp) 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>

Copilot

Pull Request Overview

This PR addresses multiple critical issues to make the image generation MCP server production-ready, focusing on fixing failing tests, updating environment variable formats, enhancing provider support, and improving development tooling.

Key Changes

Environment variables updated to provider-namespaced format (PROVIDERS__*__*)
Switch from Gemini Generative AI API to Vertex AI with support for Imagen-4
Fixed all failing tests and linting issues across the test suite

Reviewed Changes

Copilot reviewed 41 out of 50 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
tests/unit/test_utils.py	Fixed imports, added image format detection tests, improved test structure
tests/unit/test_tools.py	Complete rewrite with proper mocking and updated to new provider system
tests/unit/test_storage.py	Added helper functions for larger test images, fixed async patterns
tests/unit/test_config.py	Updated environment variable format and improved validation tests
tests/integration/test_server_integration.py	Fixed integration tests with proper provider mocking
tests/conftest.py	Added helper functions and improved fixture organization
start-mcp.sh	New script for Claude Desktop integration with environment loading
scripts/dev.py	Updated to use new environment variable format
run.sh	Updated environment variable references
pyproject.toml	Added new dependencies and proper package configuration
gpt_image_mcp/utils/validators.py	Added image format detection with magic number signatures
gpt_image_mcp/utils/path_utils.py	Improved formatting and type hints
gpt_image_mcp/utils/openai_client.py	Enhanced error handling and cost estimation
gpt_image_mcp/utils/cache.py	Improved type hints and statistics handling
gpt_image_mcp/types/models.py	Simplified field definitions and improved formatting
gpt_image_mcp/types/enums.py	Added BLACK background type, improved descriptions
gpt_image_mcp/types/init.py	Reordered imports for consistency
gpt_image_mcp/tools/image_generation.py	Major refactor for new provider system with better error handling
gpt_image_mcp/tools/image_editing.py	Enhanced validation and improved OpenAI client integration
gpt_image_mcp/tools/init.py	Reordered imports for consistency
gpt_image_mcp/storage/manager.py	Added comprehensive storage management with cleanup and compatibility methods

tests/unit/test_tools.py

gpt_image_mcp/utils/validators.py

gpt_image_mcp/storage/manager.py

…nd test patterns - Use specific Union[bytes, str] type instead of Any for image_data parameter - Update deprecation notice with specific version and date - Simplify WebP format detection return statement for better readability - Replace __new__ pattern with direct static method testing in tests 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>

rlorenzo · 2025-08-27T01:06:47Z

@lansespirit, is there any chance these changes can be merged in? I have been using this project in an app recently, and it has been working out great. I hope that others will find it helpful as well.

lansespirit · 2025-08-27T05:16:00Z

@rlorenzo Thanks for your work, good job.

rlorenzo and others added 2 commits August 3, 2025 09:39

Fixed failing tests and linting issues

68b48a6

- Add .gitignore file with Python, IDE, and build artifacts exclusions - Improve code formatting and linting compliance across all modules - Fixed failing tests - Removing API key logging

Copilot AI review requested due to automatic review settings August 4, 2025 01:49