Show HN: Tenure – Traceable AI memory with configurable memory modes (tenureai.dev)

A new tool called Tenure has been introduced as a solution for managing AI memory, emphasizing the need for configurable memory modes and explicit memory boundaries. Unlike traditional AI systems that rely on context windows and probabilistic recall, Tenure enables deterministic knowledge with versioned beliefs, clear provenance, and hard scope isolation. This new approach seeks to mitigate issues like memory drift, stale knowledge, and context bleed, which can diminish the quality of AI outputs and lead to cross-project contamination. Tenure's architecture allows for real-time logging of every decision made by the AI, ensuring full traceability and auditability. Users can oversee the AI's learning process without initiating changes, making it easy to review and manage beliefs captured during sessions. By maintaining hard boundaries between different projects and clients, Tenure aims to provide engineers with greater control over the AI's state, allowing for filtered integration of relevant knowledge while minimizing latency and costs associated with irrelevant information. This tool positions itself not just as a plugin for existing IDEs, like VS Code, but as a dedicated state management layer that can be integrated across various platforms, enhancing the reliability and efficiency of AI systems.