claude-mem and claude-self-reflect
These are competitors addressing the same problem through different approaches: one uses AI-driven compression to maintain context across coding sessions, while the other implements a simpler reflection mechanism for memory persistence, but both aim to overcome Claude's context window limitations.
About claude-mem
thedotmack/claude-mem
A Claude Code plugin that automatically captures everything Claude does during your coding sessions, compresses it with AI (using Claude's agent-sdk), and injects relevant context back into future sessions.
Implements a persistent memory system with progressive disclosure that layers retrieved context while displaying token costs, and provides skill-based search capabilities alongside a web UI viewer running on localhost:37777. Built on Claude's agent-sdk with automatic tool observation capture, semantic summarization, and support for privacy controls via `
About claude-self-reflect
ramakay/claude-self-reflect
Claude forgets everything. This fixes that. 🔗 www.npmjs.com/package/claude-self-reflect
Maintains persistent searchable conversation memory across sessions through local SQLite storage with optional cloud embeddings via Voyage AI for semantic search. Integrates with Claude Code and the Model Context Protocol (MCP) as a Docker-containerized service, automatically capturing and indexing new conversations in real-time. Enables Ralph loop workflows to share learnings across iterations and projects via pre-compaction backups and anti-pattern injection.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work