All Topics
All Topics
Technology
Technology
Design
Design
Programming
Programming
Science
Science
News
News
Gaming
Gaming
Entertainment
Entertainment
Business
Business
Finance
Finance
Sports
Sports
Health
Health
Food
Food
Travel
Travel
Art
Art
Music
Music
Books
Books
Education
Education
Politics
Politics
Personal
Personal
No algorithm. No AI slop. No ads. Just RSS. Pro-human. Indie writers. Real journalism. Open web. Chronological. Hand toasted.

MCP Server Tool Definitions Consume Excessive Context Tokens; Apideck CLI Offers Efficient Alternative

By

gertjandewilde

2mo ago· 16 min readenInsight

Summary

The article discusses a significant problem with MCP (Model Context Protocol) servers where tool definitions consume massive amounts of token context (55,000+ tokens) before any user message processing begins. This severely limits AI agent capabilities by using up valuable context window space. The author presents an alternative solution called Apideck CLI, which replaces tens of thousands of tokens of schema with an ~80-token agent prompt. This approach uses progressive disclosure via `--help` commands and structural safety baked into the binary, allowing any agent that can run shell commands to use it without requiring protocol support.

Key quotes

· 4 pulled
Before your agent has read a single user message, 55,000 tokens of tool definitions are sitting in the context window. That's over a quarter of Claude's 200k limit. Gone.
Each MCP tool costs 550-1,400 tokens for its name, description, JSON schema, field descriptions, enums, and system instructions.
We built the Apideck CLI as an AI-agent interface instead: an ~80-token agent prompt replaces tens of thousands of tokens of schema, with progressive disclosure via `--help` and structural safety baked into the binary.
Any agent that can run shell commands can use it. No protocol support required.
Snippet from the RSS feed
TL;DR: MCP tool definitions can burn 55,000+ tokens before an agent processes a single user message. We built the Apideck CLI as an AI-agent interface instead:an ~80-token agent prompt replaces tens of thousands of tokens of schema, with progressive discl

You might also wanna read