Lore Context

A control plane for AI agent memory that sits on top of your existing memory backend rather than replacing it. Exposes MCP tools for context queries that compose memory, web search, repository graphs, and tool traces into a single graded block with provenance. Runs recall and precision evals against your own datasets, routes governance reviews through a six-state lifecycle with audit logs, and exports memory as portable JSON you can migrate between systems. Ships with a repository indexing engine that surfaces files, symbols, and call graphs as read-only tools. Integrates first-class with agentmemory and provides clean adapter contracts for other runtimes. Useful when you need to prove what an agent remembered and why, share trustable context across multiple agents, or run compliance-driven deployments that require local hosting and audit trails.