Most MCP servers are empty
Most MCP servers are empty: 856 tools, 103 servers, 97% with a critical smell, and how the domain knowledge gets inside the tool description itself.
- Niveau
- Advanced
- Taal
- English
- MCP
- Metadata
- Agentic Engineering
Waar ik deze talk gaf
- AGNTCon + MCPCon Europe 2026Amsterdam, The Netherlands17 sep 2026 - 18 sep 2026Sessiepagina →
AGNTCon + MCPCon Europe 2026 in Amsterdam, 18 sep 2026
Abstract
Dit abstract staat in het Engels omdat het de tekst is die de programmacommissie heeft geaccepteerd. De talk zelf is Engelstalig.
De cijfers in het geaccepteerde abstract beschrijven het platform op het moment van inzending. Nu: 16 servers, 103 tools, 105 actieve gebruikers.
An analysis of 856 tools across 103 MCP servers found 97% of tool descriptions carry a critical smell. But smells aren't the real problem: the data is fine, what's broken is meaning. MCP isn't dead, most servers are empty, calling every endpoint and understanding nothing.
I run nine production MCP servers (50+ tools) at a 300-400 person Dutch HVAC company, used daily by non-developers. A maturity ladder emerged: most just wrap an API; a rare few teach the agent the domain, with the knowledge inside the tool description itself. The same data behind a wrapper versus a self-teaching tool is a different product.
To climb the ladder as you build: scaffold from the API docs, ship something simple, then run Introspective Context Engineering, a loop of Examine, Flag, Validate, Encode, Iterate. An AI explores the data and flags patterns by confidence; a domain expert confirms or kills the uncertain ones in an afternoon. Then telemetry takes over: every call carries a queryIntent, so the logs reveal what the agent thought it was doing, exposing the gap. Fixes take hours, not design cycles.
You'll leave with the maturity model, the discovery method, and a runnable open-source example.