AgentRecall

AgentRecall is a learning system that bridges the gap between human thinking and AI agent behavior. It provides persistent, compounding memory with automatic correction capture via MCP server, SDK, and CLI. Every mistake is recorded once and never repeated, and token savings compound over sessions.

AgentRecall sits in the Memory & Context category, so it is usually better judged by workflow fit, deployment style, and team needs than by isolated features alone. Based on the available metadata, it leans most clearly toward Persistent, compounding memory with 200-line awareness cap. A representative use case is Scattered human: structurize non-linear instructions across sessions. The listed license is MIT. GitHub stars are currently 371. Relevant tags include agent-memory, ai-agents.