Skip to main content
Glama
by LeGenAI
CHANGELOG.md1.99 kB
# Changelog ## [2.2.0] - 2025-01-05 ### 🐛 Bug Fixes - **Fixed vector dimension mismatch error**: Resolved "different vector dimensions 1536 and 3072" error - **Unified embedding models**: All knowledge bases now use `text-embedding-3-small` with consistent 1536 dimensions - **Improved search stability**: Eliminated "Hybrid search error: Invalid API key" issues - **Database consistency**: Ensured all embedding vectors use the same model and dimensions ### 🔧 Technical Improvements - Standardized embedding configuration across all files: - `supabase-knowledge-base.ts` - `advanced-knowledge-base.ts` - `knowledge-base.ts` - `smithery-index.ts` - Enhanced error handling for vector search operations - Improved database initialization and validation ### ✅ Verified Functionality - Hamming code search now returns proper results - Linear code generator matrix queries work correctly - No more API key or dimension mismatch errors - All MCP tools (`search_magma`, `get_magma_example`, `explain_magma_code`) function properly ### 📚 Documentation - Updated deployment success documentation - Added troubleshooting guide for embedding dimension issues - Improved setup instructions ## [2.1.0] - 2025-01-05 ### 🚀 New Features - Advanced hybrid search with BM25 + vector similarity - Function-specific search capabilities - Quality benchmarking system - Enhanced query expansion for mathematical terms ### 🔧 Improvements - Better text chunking for MAGMA code preservation - Improved metadata extraction - Enhanced error handling ## [2.0.0] - 2025-01-05 ### 🎯 Major Release - Complete rewrite with hybrid search architecture - 4x improvement in search relevance (20% → 84.7%) - Production-ready quality for mathematical queries - Comprehensive function indexing (4441 MAGMA functions) ## [1.0.0] - 2024-12-01 ### 🎉 Initial Release - Basic vector search functionality - PDF document indexing - MCP server implementation - Claude Desktop integration

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/LeGenAI/mcp-magma-handbook'

If you have feedback or need assistance with the MCP directory API, please join our Discord server