Learning Path
Navigate the AXO curriculum
Applications
Application and Service Optimization
Build APIs, services, and applications that agents can interact with securely and effectively.
Agents don’t just read your content — they arrive to act on behalf of a person who delegated a task. These pages cover what that requires: how to tell delegated agents from adversarial bots, what an agent needs from your API to recover from failure, the economics of agent traffic, and the protocols to build on in 2026.
Agents Are a Medium
Delegated agents vs. adversarial bots, what blanket blocking costs, and how the legal picture shifted in 2026.
The Delegation Cycle
Delegate, attempt, feedback — and what each stage needs: discovery, legible errors, idempotency, progress.
Agent–Service Relationships
The crawl-for-traffic bargain broke. Crawl-to-refer ratios, the 2026 attempts to reprice access, and what to want from agent traffic.
Principles for Agent-Accessible Services
Human accountability, capability discovery, least privilege, safe retries, legible failure, observability.
Implementation Patterns
MCP servers after the stateless revision, WebMCP, agent-friendly HTTP APIs, and Web Bot Auth verification.
Patterns in Public Work
Machine-readable docs, first-party MCP servers, the agentic checkout retreat, and prompt injection as the constraint.
Stay Updated
Analysis of AI search, crawler policy and agent standards — sent when there is something worth reading, roughly twice a month. Unsubscribe anytime.