doobidoo/mcp-memory-service v10.3.0

v10.3.0 - SQL-Level Filtering Optimization

on GitHub

🚀 Performance Breakthrough: SQL-Level Filtering

This release delivers dramatic performance improvements for large datasets through SQL-level filtering optimization, addressing issue #374.

🎯 Key Highlights

⚡ 115x Performance Speedup

• Tag filtering: 116ms → 1ms (115x faster at 1,000 memories)
• Time range filtering: 36ms → 0.49ms (74x faster)
• Memory usage: 147MB → 2.5MB (98% reduction at 10,000 memories)

🚀 New Backend Methods

• delete_by_tags: Efficient bulk deletion by tags
• get_memories_by_time_range: Time-based memory retrieval
• Complete performance benchmark suite

🔧 Enhanced API Consistency

• Standardized delete_by_tags signature across all backends
• Returns 3-tuple (count, message, deleted_hashes) for audit trail
• Improved exception handling and validation

📊 Performance Benchmarks

Tag Filtering

Time Range Filtering

Memory Usage

📦 Installation

pip install --upgrade mcp-memory-service

🔗 Related

• Closes #374
• PR #393
• Contributor: @isiahw1

📝 Full Changelog

See CHANGELOG.md for complete details.

Note: PyPI publishing is handled automatically by GitHub Actions workflow "Publish and Test (Tags)". The package will be available on PyPI within a few minutes.