Hi all, here are today's three links:

Agent-native Architectures

Dan Shipper with an excellent article on the future of applications:

The surprising discovery: A really good coding agent is actually a really good general-purpose agent. The same architecture that lets Claude Code refactor a codebase can let an agent organize your files, manage your reading list, or automate your workflows.

Agent design patterns

Lance Martin discussing the approaches that make today's AI agents effective:

A few patterns have emerged. Give agents a computer, and push actions from the tool calling layer to the computer. Use the filesystem to offload context and for progressive disclosure. Use sub-agents to isolate context. Evolve context over time to learn memories or skills. Cache to save cost / latency.

Universal Commerce Protocol

Google launches a new open standard for agentic commerce:

UCP establishes a common language for agents and systems to operate together across consumer surfaces, businesses and payment providers. So instead of requiring unique connections for every individual agent, UCP enables all agents to interact easily.

If you have any agent-related links you'd like to feature, let me know!

Cheers,
Ivan