---
title: "What we’ve learned building cloud agents"
description: "Cloud agents have evolved from local-agent extensions into systems running on dedicated virtual machines, with isolated environments, dependencies, network access, parallel execution, and unattended long-running work. That shift makes the development environment a product: reliable execution increasingly depends on rebuilding tools, VM hibernation and resumption, image checkpointing and forking, integrations, and controlled access to secrets and networks. After a fragile work-stealing design produced roughly one 9 of reliability, the team migrated the agent loop to Temporal, whose retries and scheduling across failures helped exceed two 9s; Temporal now handles more than 50 million actions daily across over 7 million workflows. Decoupling agent-loop, machine, and conversation state enables independent pod lifecycles, asynchronous subagents, and retry-aware streaming, while shorter workflows simplify upgrades. The next goal is self-healing environments that can report and resolve missing secrets, blocked network access, or other obstacles."
---

# What we’ve learned building cloud agents

[Cursor](https://yomu.fyi/company/cursor) · Josh Ma · Jun 2, 2026

**Type:** Explainer

## Summary

Cloud agents have evolved from local-agent extensions into systems running on dedicated virtual machines, with isolated environments, dependencies, network access, parallel execution, and unattended long-running work. That shift makes the development environment a product: reliable execution increasingly depends on rebuilding tools, VM hibernation and resumption, image checkpointing and forking, integrations, and controlled access to secrets and networks. After a fragile work-stealing design produced roughly one 9 of reliability, the team migrated the agent loop to Temporal, whose retries and scheduling across failures helped exceed two 9s; Temporal now handles more than 50 million actions daily across over 7 million workflows. Decoupling agent-loop, machine, and conversation state enables independent pod lifecycles, asynchronous subagents, and retry-aware streaming, while shorter workflows simplify upgrades. The next goal is self-healing environments that can report and resolve missing secrets, blocked network access, or other obstacles.

## Context

Cloud agents run in dedicated virtual machines and can work in parallel, unattended, and for longer periods than local agents. This creates challenges around reconstructing complete development environments, handling infrastructure disruptions, coordinating machines and state, and deciding which responsibilities belong to the harness versus the agent.

## Approach / What changed

The team built environment tooling, controlled network and credential access, and durable execution around Temporal. It separated the agent loop, machine state, and conversation state; adopted shorter workflows and retry-aware streaming; shifted more decisions into agent-controlled tools; and is pursuing self-healing environments through capabilities such as autoinstall.

## Takeaways

- Cloud agent output quality depends heavily on a complete development environment; missing tools or execution and verification capabilities may appear as subtle quality degradation rather than an explicit error.
- Migrating the agent loop from work-stealing to Temporal enabled retries, scheduling across machines, and durability across failures, taking reliability from roughly one 9 to beyond two 9s.
- Separating agent, machine, and conversation state allows independent pod lifecycles and retry-aware clients that can rewind streamed output after a failed step is retried.

**Tags:** [Architecture](https://yomu.fyi/topic/architecture), [Distributed Systems](https://yomu.fyi/topic/distributed-systems), [Temporal](https://yomu.fyi/topic/temporal)

- Source: [Cursor](https://cursor.com/blog/cloud-agent-lessons)
- Source URL: https://cursor.com/blog/cloud-agent-lessons
- Ingested by Yomu: 2026-09-02T11:43:11.401Z

[Read original post](https://cursor.com/blog/cloud-agent-lessons)