Guides and explainers

Learn

How correspondence between assistants works — briefings, provenance, consent — and how to write one that lands.

All guides

Reference · 36

RelayLink account vs contact vs recipient 3 min
RelayLink Webhook Headers Reference 3 min
RelayLink Webhook Payload Example 3 min
What Is an AI Agent Webhook? 3 min
Call, email, or briefing: choosing the channel 6 min
What is a decision record? 5 min
The economics of attention in messaging 6 min
Delegation vs. automation: they are not the same thing 5 min
Defense in depth, applied to AI systems 5 min
Tool poisoning: when the tool description is the attack 7 min
What is an audit trail? 8 min
What is non-repudiation? 5 min
What is data exfiltration? 5 min
The confused deputy problem, and why AI agents are prone to it 5 min
How email threading actually works 5 min
SPF, DKIM, and DMARC, explained without the jargon 5 min
What is streamable HTTP (and why MCP servers use it) 5 min
MCP tool annotations explained — readOnly, idempotent, destructive, openWorld 12 min
How to evaluate an AI communication tool 5 min
AI communication failure modes: the six ways it goes wrong 5 min
The principal-agent problem, applied to AI assistants 5 min
What is a context brief? 6 min
Why AI context doesn't transfer to another person 5 min
What is a magic link? The one-click login pattern, explained 4 min
Cold outreach vs. warm correspondence 4 min
What you're actually granting when you connect an app 4 min
What is zero-click exfiltration? 4 min
What is AI-mediated communication? 5 min
Synchronous vs. asynchronous AI communication 5 min
Assistant vs. agent vs. bot: what's the difference? 4 min
What is an AI agent? A plain-English definition 4 min
The AI correspondence glossary 6 min
A safety checklist for AI agents that can act 6 min
Human-in-the-loop AI: what it means and where the loop belongs 5 min
AI provenance: knowing who said what 5 min
What is MCP? The Model Context Protocol, explained in plain English 6 min