The Age of Async Agents — Cognition's Walden Yan & OpenInspect's Cole Murray

Autonomous background agents are transforming software engineering by enabling systems to move from simple model-wrapping to end-to-end task execution, such as generating pull requests from specifications. A critical architectural decision involves placing the agent "in-the-box" for simplicity or "out-of-the-box" for enhanced security and state management. Successful implementation requires robust infrastructure, including efficient repository setup and specialized file systems that handle state persistence without massive overhead. While multi-agent swarms remain an aspirational goal, current practical applications prioritize single-agent workflows that handle SRE triage, customer support, and code maintenance. As these systems scale, managing memory, ensuring code quality through strict linting, and establishing clear boundaries between modules become essential to prevent codebase regression and maintain long-term system reliability.