Side-by-side comparison of two Model Context Protocol servers — install paths, tools, quality signals, and directory engagement so you can pick the right one for Claude, Cursor, and other MCP clients.
Local-first persistent memory for AI agents. SQLite + local embeddings (all-MiniLM-L6-v2), hybrid semantic + FTS5 recall, secret detection, and contradiction handling. 6 MCP tools (remember, recall, forget, feedback, context, status) over stdio. Zero cloud, no API keys, fully offline. Works with Claude Desktop/Code, Cursor, and Windsurf. npm install -g @hbarefoot/engram
Loads your personal profile, mined from your local Claude Code/Codex/OpenCode logs, so your agent works like you instead of a cold start. One tool, loademuloprofile. Run with uvx emulo mcp.
Quality signal
53/100 (Fair)
55/100 (Good)
Install path
npx · high
npx · high
Engagement
0 0 0 7
0 0 0 216
Tools
Local SQLite database with all-MiniLM-L6-v2 embeddingsHybrid semantic and FTS5 full-text search recallAutomatic secret detection on every write operationSix MCP-native tools: remember, recall, forget, feedback, context, statusFully offline operation with zero external dependenciesIn-process integration with no separate server or IPC overhead
Mines local AI coding session logs without uploading raw dataSupports multiple AI coding platforms including Claude Code, Codex, Copilot CLI, OpenCodeCreates layered profiles for task-specific agent behaviorIntegrates natively or through adapters with various AI agentsOperates locally with zero dependencies and MIT license
