DigitalOcean Managed Agents Brings Managed Cloud Infrastructure to AI Agents

DigitalOcean recently launched DigitalOcean Managed Agents in public preview, according to InfoQ. The offering is described as a managed cloud infrastructure layer for AI agents, with isolated microVM runtimes, governed tool access, and serverless AI inference, in a report by Sergio De Simone.

According to DigitalOcean, agentic workflows differ substantially from traditional cloud-based applications, and developers running agents on conventional VMs must build and manage the supporting runtime environment themselves. DigitalOcean said developers are forced to invest in plumbing work to preserve an agent's context, persist artifacts and keep them accessible beyond the agent that created them, coordinate parallel work, and security-harden access to tools. Further challenges cited include maintaining spare VM capacity so agents can start quickly, and provisioning and configuring compute resources on demand.

DigitalOcean Managed Agents aim to address these challenges through two integrated services: a Harness Runtime and an Action Gateway. Built on a lightweight microVM, the Harness Runtime provides persistent, isolated compute environments across supported harnesses, including coding agents such as Claude Code, Codex CLI, and OpenCode, general-purpose agents such as Hermes, and custom agents built with LangGraph. DigitalOcean said the runtime can persist conversational history and working state across sessions, which can be paused, resumed, or forked, and can automatically pause agents when idle, meaning no ongoing LLM or tool calls. Sessions run on security-hardened compute and storage and can be launched in parallel across repositories and tasks.

The Action Gateway is the agent's interface to external tools and services through a unified MCP endpoint, with more than 16,000 tools currently available, including Web Search, Web Fetch, Browser Automation, DigitalOcean infrastructure management APIs, and connectors for platforms such as GitHub, HubSpot, and Stripe. DigitalOcean said the gateway implements centralized permission management, human approval for sensitive operations, and rate limiting, retries, backoff, and timeouts.

Commenting on X.com, Elvis Saravia, a former Meta AI engineer now building Dair.ai, wrote that he liked the focus on performance and cost reduction, saying many other agent management platforms are cost-prohibitive and make it hard to scale agents in production. Seaotter platform architect and founder Ryan Martin noted that "Pause-when-idle + governed tools is the ops pattern that scales."

InfoQ noted DigitalOcean Managed Agents arrive around the same time as Docker's Cloud Sandboxes, with overlap at the sandbox and runtime level. Both offer microVM isolation and persistency, but InfoQ said Docker Cloud Sandboxes focus more on developer workflows, particularly moving a sandbox between local and cloud compute, while DigitalOcean Managed Agents focus more on production infrastructure with more orchestration and tool support.