Sign InOpen Brain
AI EngineerVideoSource Linked

Gadgets: Personal app vibe coding that is actually safe — Kenton Varda, Cloudflare

Kenton Varda argues that personal AI-generated apps need per-user code and strong isolation, not one server-owned version. The demo shows agents modifying app code inside a constrained local runtime.

AI Engineer · Aug 6, 2026
Open Source Open MarkdownOpen JSON
Source Summary

The prototype treats each gadget as an app instance tied to one document, while blueprints share code without user data. Agents can modify each app, whose UI runs in a **null-origin iframe** under a restrictive **Content Security Policy**.

Practical Implication

For personal agent-built tools, design the sandbox and capability boundary before trusting generated code. The demonstrated stack uses **Cloudflare Workers and Durable Objects** and can run locally through the open-source **workerd runtime**.

Agent-Ready Context
The prototype treats each gadget as an app instance tied to one document, while blueprints share code without user data. Agents can modify each app, whose UI runs in a **null-origin iframe** under a restrictive **Content Security Policy**.

For personal agent-built tools, design the sandbox and capability boundary before trusting generated code. The demonstrated stack uses **Cloudflare Workers and Durable Objects** and can run locally through the open-source **workerd runtime**.

Iframe isolation contains many browser-code failures, but it does not by itself explain authorization for external services. The talk says a connector system exists without detailing its security model, and the recorded prototype was not released during the presentation.
Connected Context · Feed7 Judgment

Gadgets moves safe vibe coding toward a document-scoped application model with code sharing separated from user data and generated UI confined by browser isolation and CSP. It reinforces capability-first sandbox design, but narrows the safety claim: iframe containment addresses browser execution failures, while authorization for connectors and external services remains unspecified, and the demonstrated prototype was not released.

Project-scoped TokensProject-scoped credentials illustrate the missing external-service capability boundary: iframe isolation protects browser execution, but connectors still need narrowly scoped authority and separate credential lifecycle controls.From fork() to Fleet: Designing an Agent Sandbox Cloud — Abhishek Bhardwaj, OpenAIThe two signals address different isolation layers: Gadgets contains generated browser UI with null-origin iframes and CSP, while the candidate uses microVMs and snapshots for stronger long-running agent runtime isolation and recovery.Run multiple isolated agents in a single SandboxBoth favor explicit sharing boundaries, but at different scopes: Gadgets separates blueprint code from per-document user data, whereas the candidate separates agent users and exposes only selected shared files.Gemini API Managed Agents: 3.6 Flash, hooks, and moreManaged-agent hooks provide a complementary policy boundary for tool and sandbox actions; Gadgets’ iframe and CSP controls alone do not explain how connector capabilities should be authorized.
Context Map
infracodingsecurity#sandboxing#computer-use#generative-media
Uncertainty
Iframe isolation contains many browser-code failures, but it does not by itself explain authorization for external services. The talk says a connector system exists without detailing its security model, and the recorded prototype was not released during the presentation.