Show HN: Open-source control engine for agents (attestorportal.com)

A new open-source control engine called "Attestor" has been introduced, aimed at enhancing the security and accountability of AI-driven operations. Positioned as a mediator between AI-generated actions and the execution systems, Attestor ensures that any action proposed by an AI must undergo rigorous checks for authority, scope, evidence, and proof before it can be executed. For example, while the engine can generate a refund request, the actual execution is contingent upon verified approvals, making it clear that intent alone does not equate to permission. This development is significant for the AI/ML community as it addresses crucial safety concerns surrounding automated decision-making systems. By preventing unauthorized or unsafe actions from being executed, Attestor minimizes the risks associated with malicious inputs or oversight errors. The engine meticulously evaluates proposals, considering factors such as approval state, evidence sufficiency, and compliance with preset limits, ultimately leading to transparent and accountable decision-making processes in AI operations. In essence, Attestor strives to provide a robust framework that guarantees actions taken by AI systems meet stringent trustworthiness criteria.