Back to browse
GitHub Repository

Three-layer memory architecture for AI coding agents. Persistent facts, cross-session task tracking, and ephemeral execution — so agents stop forgetting.

2 starsJavaScript

Mneme–Persistent memory for AI agents without vector search or RAG

by xqli·Feb 26, 2026·1 point·4 comments

AI Analysis

●●SolidBig BrainShip It

Structured memory layers for agents—but vector search already solves this problem.

Strengths
  • Three-tier separation (long/mid/short-term) is genuinely clever architecture for agent persistence
  • Dependency-aware task tracking avoids re-running solved work across sessions
  • No RAG required; pure structured storage means predictable, auditable memory
Weaknesses
  • Competes against established RAG/vector solutions (Pinecone, Weaviate) without clear advantage
  • Documentation is thin; unclear how this outperforms prompt engineering + session logs
Category
Target Audience

AI agent builders, OpenCode users, autonomous coding systems

Similar To

LangChain · Mem0

Similar Projects

AI/ML●●Solid

Cecil – open-source memory and identity protocol for AI

Clever multi-layer memory architecture (seed/narrative/delta), but "AI memory" is well-explored territory.

Big BrainZero to One
JohnKnopf
113mo ago
AI/ML●●Solid

Synrix local-first memory engine (O(k) retrieval, no vectors, no cloud)

Swapping global vector scans for O(k) prefix/deterministic retrieval is a clever pivot that could cut latency and cost for local agent memory. The repo ships a usable Windows binary plus an MIT Python SDK and LangChain-friendly badges — enough to test the claim quickly — but the core engine is proprietary and lacks reproducible benchmarks, so you’ll want evidence before trusting it at scale.

Niche GemWizardry
JosephjackJR
204mo ago
AI/ML●●Solid

Embenx – agentic memory layer for AI agents

MCP-native memory with synthetic data generation for AI agent retrieval workflows.

Ship ItNiche Gem
akarnam37
102mo ago