"MCP Was a Bad Idea": Devs Argue Smarter Models Made Most MCP Servers Obsolete
udmrzn · x · 2026-09-28
A widely shared critique titled "Why MCP Was Always a Bad Idea" argues that MCP — the protocol Anthropic released in November 2024 — is a relic from an era when models couldn't reason or act autonomously.
Key points:
- MCP adoption exploded, but each server's tool schemas bloated context, spawning an ecosystem of wrappers and search/execute layers from Composio, MintMCP, and Pipedream — a veritable "MCP industrial complex."
- Meanwhile models gained strong reasoning and code-execution abilities: modern LLMs write their own scripts, call unfamiliar APIs directly, and learn CLIs by reading --help.
- Thin MCP servers that merely wrap an API or CLI are now unnecessary; terminal-capable agents hitting HTTP APIs directly are far more flexible.
- The author recommends relying on existing standards (e.g., Accept: text/markdown) instead of bespoke protocols, and suggests most MCP servers can simply be deleted.
More from coding & agent
- AI agent digs into game code, records its own footage to cut a promo trailer — AIandDesign · 2026-09-28
- JevControl open-sources a tool to test if smaller models can run your agent's routine choices — airesearch12 · 2026-09-28
- Claude Code CLI connects to Gmail easier than Google's own Antigravity, dev finds — randal_olson · 2026-09-28
- Pelican test on MiMo 2.6: plan mode cuts agent calls to 12-17 vs heavy iteration — PilgrimofHaqq2 · 2026-09-28
- Memory history is not reasoning history: agents lack a causal provenance layer — puppy_lover_2021 · 2026-09-28
- mark_k: AI coding is approaching a step change — systems only AI can debug — mark_k · 2026-09-28