Please Don't Kill My Vibe: Empowering Agents with Data Flow Control

TL;DR

This paper advocates for native Data Flow Controls in systems to manage undesirable data flows in LLM agents, aiming to improve security, policy enforcement, and process integrity.

Contribution

It introduces the concept of Data Flow Controls for agent ecosystems and presents initial work on implementing DFC in database management systems.

Findings

Developed a portable DFC prototype for DBMSes

Identified key challenges in managing data flows in agent systems

Outlined a research agenda for DFC in agent ecosystems

Abstract

The promise of Large Language Model (LLM) agents is to perform complex, stateful tasks. This promise is stunted by significant risks - policy violations, process corruption, and security flaws - that stem from the lack of visibility and mechanisms to manage undesirable data flows produced by agent actions. Today, agent workflows are responsible for enforcing these policies in ad hoc ways. Just as data validation and access controls shifted from the application to the DBMS, freeing application developers from these concerns, we argue that systems should support Data Flow Controls (DFCs) and enforce DFC policies natively. This paper describes early work developing a portable instance of DFC for DBMSes and outlines a broader research agenda toward DFC for agent ecosystems.