BlockEdge

Technology and digital life

DigitalOcean previews managed cloud infrastructure for AI agents

Originally reported November 16, 2026. DigitalOcean has launched a public preview of DigitalOcean Managed Agents, a service intended to take over parts of the infrastructure work involved in running AI agents. It combines two services: Harness Runtime for agent execution and Action Gateway for access to external tools.

Harness Runtime uses isolated microVMs to run supported agents, including Claude Code, Codex CLI, OpenCode and Hermes, as well as custom agents built with LangGraph. According to DigitalOcean, it preserves conversation history and work state across sessions. Sessions can be paused, resumed or forked, and automatically pause when idle—when no LLM or tool call is in progress. Developers can also connect agents to internal services without exposing those services to the public internet.

Action Gateway provides a unified MCP endpoint for external tools. The article reports that more than 16,000 tools are available, spanning web search, browser automation, DigitalOcean infrastructure APIs and connectors for services such as GitHub, HubSpot and Stripe. Its described controls include per-agent permissions, mandatory human approval for sensitive actions, rate limits, retries, backoff and timeouts.

The article compares the preview with Docker Cloud Sandboxes, noting that both offer microVM isolation and persistence. It characterizes Docker’s offering as more focused on developer workflows and movement between local and cloud compute, while DigitalOcean’s emphasizes production infrastructure, orchestration and tool access. These are descriptions of the services, not independently verified performance or cost results.

Source: infoq.cn (originally published 2026-11-16 16:42 UTC+8)