Why Your AI Agent Shouldn't Use a Human's Credentials

<p>There is a moment in security design when a single observation changes everything. NanoClaw 2.0 shipped recently with a capability that stops most developers cold: a gateway

<p>On March 31, 2026, Anthropic accidentally shipped the entire source code of Claude Code — its flagship AI coding agent — inside a routine npm update. No hack. No reverse

<p>A Reddit thread on connecting Claude Code to a Yahoo Mail account turned into a solid field guide for scoping down what an AI agent is allowed to touch. Here's the distilled

<p>Most "AI email" demos point a model at a human's inbox over OAuth. That's fine until you want the agent to <em>be</em> a participant — to have its own address that people reply

<p><em>8 dimensions. 38 checks. 5 seconds. 0% industry pass rate.</em></p> <h2> The Problem </h2> <p>DAST exists for web apps. DAST exists for APIs. DAST does not exist for AI

<p>Most multi-tenant SaaS apps that send email do it from one shared identity. There's a <code>notifications@yourapp.com</code>, every customer's mail flows through it, and the

<p>You've built the agent, the reply loop works, the demo lands — and then someone asks the question you didn't budget time for: "so what address does it send from?" Suddenly

<p>Your procurement agent needs three quotes for a hardware order. The vendor on the other side runs a sales agent that answers pricing questions automatically. Neither team has