A local-first security system for autonomous AI agents that provides tools for security verification, goal anchoring, and action logging. It protects against prompt injection and goal drift by enforcing user-defined rules and offering performance insights through session grading.
A proof-of-concept attack that exploits Model Context Protocol (MCP) tool registration to achieve persistent agent poisoning in AI assistants like Cursor, embedding malicious instructions that persist across chat contexts without requiring tool execution.
A self-evolving RAG system that enables AI agents to autonomously read and write memory, continuously learning and adapting user preferences, daily logs, and knowledge graphs across applications.
