gertjandewilde

We built a unified API with a large surface area and ran into a problem when building our MCP server: tool definitions alone burned 50,000+ tokens before the agent touched a single user message.

The fix that worked for us was giving agents a CLI instead. ~80 tokens in the system prompt, progressive discovery through --help, and permission enforcement baked into the binary rather than prompts.

The post covers the benchmarks (Scalekit's 75-run comparison showed 4-32x token overhead for MCP vs CLI), the architecture, and an honest section on where CLIs fall short (streaming, delegated auth, distribution).

Most APIs were designed for human developers, not autonomous agents. As LLMs start selecting endpoints and generating arguments directly from your schema, ambiguity and weak error semantics become production issues. This post outlines practical API design patterns that make APIs more reliable for agent-driven workflows.