Key features implemented:
- Updated .gitignore to include additional cache and coverage directories (.mypy_cache/, .pytest_cache/, coverage/, htmlcov/)
- Enhanced XSS protection in pkg-manager plugin by adding HTML escaping for all user data in page rendering functions
- Improved PL injection security in plugin-loader with enhanced static source code analysis including base64 decoding checks and string concatenation bypass detection
- Strengthened configuration file loading security using ast.literal_eval for safe parsing and stricter code execution prevention
- Added comprehensive dangerous pattern checks in PL injection static analysis to prevent various bypass techniques
The updates focus on security hardening across the plugin system, particularly addressing input sanitization and code injection vulnerabilities while expanding the project's ignore patterns for better repository cleanliness.
Key features implemented:
- Updated package metadata and dependencies in PKG-INFO, setup files
- Added main.py entry point for backward compatibility with README launch method
- Enhanced CLI with config options, system info command, and proper signal handling
- Implemented minimal PluginManager loading only plugin-loader core plugin
- Refactored PluginLoader to follow minimal core design, removed sandbox/isolation complexity
- Updated auto-dependency plugin with safer PL injection mechanism and disabled pl_injection
- Removed legacy plugin files (firewall, frp_proxy, ftp_server, multi_lang_deploy, ops_toolbox, security_gateway) as functionality moved to core plugin system
- Improved gitignore with comprehensive ignore patterns
The changes implement a minimal core framework design where only the plugin-loader is directly loaded by the core, with all other plugins managed through the PL injection mechanism, significantly simplifying the architecture.
- Added performance-optimizer plugin with FastCache LRU caching, ObjectPool for object reuse, BatchProcessor for bulk operations, MemoryArena for pre-allocated memory, PerfProfiler for low-overhead timing, and StringIntern for deduplication
- Implemented high-performance routing optimizations in router.py using @lru_cache decorators for path matching and parameter extraction functions
- Created comprehensive plugin infrastructure with manifest.json configuration and unified access interface through PerformanceOptimizerPlugin class
- Enhanced system performance through multiple optimization strategies targeting different bottlenecks with measurable performance gains from 2x to 100x improvement ratios
Key features implemented:
- Updated README.md to prominently display MIT license badge and include full license text in expandable section
- Enhanced .gitignore with comprehensive file patterns for Python, dependencies, logs, and build artifacts
- Added detailed plugin manifest example showing license declaration in plugin configuration
- Included license information in plugin development documentation section
The updates provide clear license visibility and improved development workflow configuration.
