🚀 New Essay: Agents, APIs, and the Next Layer of the Internet

mercury0100

18 Jun 2025 — 1 min read

A container port, symbolising structured routing and real-time discovery — a visual metaphor for agentic infrastructure. Image by William William via Unsplash

LLMs aren’t just answering questions anymore — they’re acting. But to act intelligently, they need tools.

Our latest article on Towards Data Science explores the future of agent-tool interaction, and the two leading approaches shaping it:

MCP (Model Context Protocol): a structured protocol for agent-to-agent calls
Invoke: an open, runtime-first model for dynamic tool use based on agents.json

We break down the architectures, trade-offs, and philosophies behind both — and explain why the future of the agentic web depends on getting this layer right.

👉 Read the full article on Towards Data Science

Keywords: agentic web, LLMs, APIs, tools, model context protocol, Invoke, AI infrastructure, open agentic systems, runtime tool discovery

You Don’t Know What You Want Until You Want It

Why Runtime Discovery Is the Future of the Agentic Web Invoke Labs | June 2025 Abstract As the agentic web emerges, one thing becomes clear: the interface between models and the world must be dynamic, lightweight, and open. While current approaches like MCP (Anthropic, 2023) attempt to standardize tool use through

How to Connect Any LLM to Any API in Seconds

Invoke Labs | June 2025 Turn your LLM into an agent that can take action — not just talk. Modern language models like GPT-4.1 are brilliant at reasoning, writing, and planning. But if they can’t take action — if they’re stuck inside a sandbox — they’re just glorified typewriters. Invoke

LLMs aren’t just answering questions anymore — they’re acting. But to act intelligently, they need tools.

Read more

You Don’t Know What You Want Until You Want It

How to Connect Any LLM to Any API in Seconds