ShipItAndPray

[](https://glama.ai/mcp/servers/ShipItAndPray/mcp-turboquant) 📇 🏠 🍎 🪟 🐧 - LLM quantization via tool call. Convert models to GGUF, GPTQ, and AWQ formats. Recommend optimal quant settings, evaluate quality, and push to Hugging Face Hub.

** First MCP server for data compression. Gives AI agents 7 tools: compress, decompress, analyze, store, retrieve, list, stats. Auto-picks best algorithm. 60x compression on docs, 30x on SQL, 20x on logs. Lossless round-trip. Zero dependencies.

[](https://glama.ai/mcp/servers/ShipItAndPray/mcp-memory) 📇 🏠 🍎 🪟 🐧 - Smart memory with exponential decay. Memories strengthen on access and fade when unused — solving the Karpathy problem of unbounded context growth. 7 tools, SQLite-based, zero dependencies.

** First MCP server for live market data. 8 tools: price, candles, order_book, market_cap, trending, analyze (RSI/SMA/z-score), compare, feargreed. Zero API keys. Zero dependencies. 12/12 evals on live data.

Compressed Persistent Memory for AI Agents — MCP server with zlib compression and SQLite storage

**GitHub:** https://github.com/ShipItAndPray/mcp-memory Solves Karpathy's complaint about LLM memory: memories decay over time, topics frequency-weighted, one-time questions don't become obsessions. 7 tools. Zero deps. **Install:** `npx mcp-memory`