How task containers give AI agents real infrastructure without idle cost

Upsun introduced task containers on August 12, 2026, as single-purpose, short-lived containers that execute one command, terminate when it completes, and incur charges only while running, providing an alternative to both isolated sandboxes and continuously running infrastructure. Defined in a project’s configuration, they can support AI agent sessions, database maintenance, exports, migrations, and application-triggered batch jobs while accessing project services and production-like preview environments that clone data, files, and services from a parent environment. Although tasks inherit platform container protections such as namespace isolation, dropped capabilities, seccomp profiles, cgroup limits, and network isolation, stricter sandboxing for untrusted code requires additional firewall, mount, variable, or bubblewrap configuration. A demonstration showed an AI-triggered task using the Anthropic API to modify a repository, open a pull request, and generate a preview environment, while a separate non-AI task exported prompt and model-output records for later evaluation. Tasks can access applications and databases but cannot receive inbound traffic, have configurable runtimes of up to one day, and allow three concurrent executions by default, with additional jobs queued. Upsun is also developing a performance-focused agent that will run as a task against observability data to produce application recommendations.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.