This edition features a recent webinar in collaboration with Emmanuel Paraskakis on how to speed up your API design process using LLMs, while keeping in mind the developer experience and business alignment. Check out our talk and the resources I provided, including a report on what LLMs are best for your API design efforts. We also look at OpenAPI generators and how to go from spec to skeleton code and SDK libraries.
Also featured this week is a look at the emerging Event and Webhook Delivery Semantics RFC. From the RFC: "Despite widespread deployment, there is no common, interoperable delivery semantics contract governing how event producers, consumers, and intermediaries signal delivery attempts, retries, acknowledgment, or failure." - check out and weigh in on this emerging RFC.
Next, we look at MCP and how it fits into the ecosystem of emerging agent skills, APIs, and the new UTCP for specifying how to invoke APIs without an MCP server layer. Finally, we look at the economics of APIs and AIs and a possible trend in APIs shuttering in the age of AI.
-- Happy Reading!
Featured Articles
Upgrade Your API Design with AI
"Rushing to code" with GenAI creates technical debt. To build production-grade APIs, you need a structured approach. James Higginbotham (LaunchAny) joins us to demonstrate the ADDR process: a methodology that turns LLMs into "API Design Coaches". You will learn to cut design time by 60% while delivering high-quality, business-aligned, and AI-ready APIs. [maven.com]
API-First with OpenAPI Generator: From Spec to REST API and Type-Safe SDKs
This guide demonstrates a contract-first API development workflow using OpenAPI Generator with Maven in Java, covering the complete lifecycle from specificat... [malkomich.github.io]
Event and Webhook Delivery Semantics
Event- and webhook-based integrations are a common application-layer mechanism for interoperability across Internet services, yet they lack a shared delivery semantics contract. Existing implementations vary widely in retry behavior, acknowledgment signaling, failure classification, idempotency identifiers, and replay handling, resulting in fragile integrations and ambiguous operational expectations. [datatracker.ietf.org]
MCP vs Agent Skills: Capabilities and Procedures Explained | Layered System
MCP and Agent Skills are complementary primitives. MCP provides capabilities (what agents can do). Skills provide procedures (how agents should approach it). [layered.dev]
MCP vs API: Why They're Not Competing
MCP vs API: Why They're Not Competing Why MCP Exists You've probably heard of Model Context Protocol (MCP) being mentioned alongside modern AI systems. But if you're wondering why it even ... by Franz Andel [medium.com]
Model Context Protocol (MCP) vs. Universal Tool Calling Protocol (UTCP)
Compare MCP and UTCP for agentic AI tool-calling, including architecture trade-offs, performance, flexibility, and ecosystem support. by J Simpson [nordicapis.com]
Business of APIs
From Cost Center to Revenue Driver: Rethinking the API and AI Mix
How AI is reshaping the API economy, turning APIs into revenue channels through usage-based pricing, governance, and agent-driven demand. by Chris Darvill [nordicapis.com]
Why Public APIs Are Shuttering in the Age of AI
Why public APIs are shutting down in the age of AI, from monetization pressures to LLM scraping, and what it means for developers. by Art Anthony [nordicapis.com]
|
|
