IAI
Local long-term memory MCP server for AI coding assistants with three-tier episodic, semantic, and procedural memory storage.
Provides persistent local long-term memory for AI coding assistants without cloud dependencies, using a three-tier storage system of episodic, semantic, and procedural memory backed by LanceDB with local embeddings. All data is encrypted with AES-256-GCM and the system claims 99%+ verbatim recall with sub-100ms retrieval latency. Runs as a Python daemon with a TypeScript MCP wrapper communicating via stdio.
Source
Repository: https://github.com/codeabra/iai-personal-memory-engine
Maintain IAI?
Let people know it's listed here — add the badge (live metrics, light/dark aware) or a plain link to your README or docs.
[](https://getagentictools.com/mcp/codeabra-iai-personal-memory-engine?ref=badge) Related mcp
Agent Recall Correction-first agent memory with a precision KPI that tracks whether agents heed warnings, across 5 local-only layers.
Proxmox Enables AI systems to manage Proxmox virtualization environments with tools for monitoring nodes, controlling virtual machines, tr
AWS Bridges AWS services with the AWS SDK v2, enabling secure access and management of AWS resources for cloud infrastructure and data
Postman Connect AI agents, assistants, and chatbots directly to your APIs on Postman.
Kubernetes Control and monitor K8s clusters for management and debugging.
Azure DevOps Integrates with Azure DevOps to enable AI-driven management of projects, work items, repositories, pull requests, branches, and pi