Holtstock-Brown proposes long-running agents for continuous vulnerability scanning
Holtstock-Brown argued that traditional security tools are being outpaced by the volume of code generated in the agentic era. He proposed replacing static solutions with autonomous agents that continuously monitor and resolve vulnerabilities.
Holtstock-Brown suggested that companies must adopt long-running agents to manage the growing security risks associated with agentic systems. He noted that it is currently easier to red-team than blue-team, creating a critical need for proactive defense mechanisms. Holtstock-Brown proposed replacing traditional security tools, such as Dependable, with agents that continuously improve an application's security posture. He described an experimental setup using a DeepSeek V4.1 Flash agent to check for potential vulnerabilities daily and pass identified issues to coding agents for resolution. This approach mirrors an Anthropic article on their code review process, which involves turning bugs into evals and enabling security loops. Holtstock-Brown argued that as code volume increases, companies must beat attackers to the punch by using agents to monitor supply chain dependencies and code changes continuously. He cited a loop where an agent checks an app daily to identify performance improvements, illustrating the potential for autonomous, ongoing optimization and security maintenance. He stated that one of the easiest wins for people would be to have a long-running agent that supersedes Dependable.