The cloud is changing#
For the last two decades, cloud computing has required humans to do the heavy lifting. You learn the platform, read the docs, write the config files, set up the pipelines, and monitor the dashboards. AWS, GCP and Azure all follow the same pattern: the cloud provides the tools, and you figure out how to use them.
The agentic cloud flips this model. Instead of you operating the cloud, an AI agent does it for you. You describe what you need (for example, “deploy my app with a CDN, database, and custom domain”) and the agent handles provisioning, configuration, DNS, SSL and monitoring on its own.
This isn’t a future concept. It’s happening now. Coding assistants like Cursor, Windsurf, and Claude can already deploy infrastructure when given the right APIs. Agents will manage cloud infrastructure. The open question is whether your cloud platform is ready for them.
Traditional vs agentic cloud#
The operational model is fundamentally different.
| Traditional cloud | Agentic cloud |
|---|---|
| Read documentation, learn CLI, configure manually | Agent discovers capabilities, deploys via natural language |
| Copy-paste API keys, manage secrets, configure SDKs | Single access code with scoped permissions and budget limits |
| Write scripts for scaling, monitoring, alerting | Agent monitors, scales and resolves issues autonomously |
| Separate dashboards for billing, resources, logs | Agent sees everything in one discovery call |
| Requires DevOps expertise and certifications | Anyone can deploy: describe what you want and the agent builds it |
How it works on Edge#
Edge Network is designed from the ground up for agents. The workflow is straightforward:
- Generate an agent access code: in the console under Agent Access, create a code with the permissions and budget your agent needs. This is a scoped credential, not your account API key.
- Give your agent the URL and code: pass your coding assistant the Edge agent endpoint and the access code. The agent calls the discovery API and immediately understands what it can do.
- Describe what you need: “Deploy my Next.js app with a CDN and custom domain” or “Set up a staging environment that mirrors production.” Natural language, not YAML.
- The agent handles the rest: provisioning, configuration, DNS, SSL and deployment, all handled autonomously. The agent reports back what it did, what it cost, and how to access your resources.
Built for agents#
Every part of the Edge API is designed with autonomous agents in mind.
- Discovery API: One endpoint returns everything an agent needs: available products, current usage, plan limits and budget. No documentation crawling.
- Agent access codes: Scoped, time-limited credentials designed for agents. Set budget caps, restrict products, and control permissions without sharing your account keys.
- Project-based resources: Group VMs, CDN deployments, DNS zones, and storage buckets into a project. Health check, scale, or tear down an entire app in one call.
- Dry run mode: Preview any action before committing. Your agent can show you exactly what would happen and what it would cost before making changes.
- Budget-aware operations: Every API response includes cost context. Hard and soft caps prevent runaway spending. Your agent operates within limits you define.
- Human-readable responses: Every response includes a
tell_userfield, a plain English summary your agent can relay directly to you without interpretation.
Who the agentic cloud is for#
- Developers: Stop context-switching between your IDE and cloud dashboards. Your coding assistant deploys while you keep building.
- Founders and vibe coders: No AWS certifications needed. Describe what you want and ship. The agent handles the infrastructure complexity you shouldn’t need to learn.
- Agencies: Deploy client projects faster. Give each project its own agent access code with budget limits. Scale without scaling your ops team.
- Enterprises: Your DevOps team becomes a platform team. Define policies and budgets, let agents handle execution. Humans focus on strategy, not tickets.
Keep learning
Edge explained4 min read
What is the edge cloud?
The cloud gave us on-demand infrastructure. The edge cloud puts that infrastructure where it actually belongs: next to your users.
Security4 min read
Verified AI agents with Web Bot Auth
The bot problem changed. Some automation is your customers' AI agents trying to buy from you. Security now means telling declared automation apart from anonymous automation and treating each accordingly.
Security5 min read
Security hardening
Step-by-step practices to harden your Edge infrastructure: authentication, firewalls, API security and agent access control.