A research investigation into federated MCP gateways. One hub process, every MCP server, every AI client. Tool schemas discovered on demand rather than preloaded. Cache, cost telemetry, and retrieval learning shared across concurrent sessions.
Each AI coding session that integrates with MCP inherits a per-session architecture designed for one-to-one client-to-server use. At cross-session scale the design produces three observable regressions.
Each AI coding session spawns its own MCP server children. With N sessions × M MCPs the process count grows as O(N·M). At five sessions and thirty-six MCPs the resident process set reaches 180 processes and roughly 9 GB of memory — before any model call has been made.
Every MCP tool schema is preloaded into each session's context window. 400+ tool definitions consume approximately 150K tokens on a 200K-context model, leaving 75% of the working set on tool prospectus rather than task content.
Independent sessions cannot share cache, cost ledgers, or retrieval signals. Every session re-computes the same lookups, re-pays the same API costs, and re-discovers the same failure modes. There is no shared substrate for learning.
A hub process owns the MCP server lifecycle. AI clients connect over HTTP and interact exclusively through three meta-tools. Tool schemas are retrieved on demand.
Measured on a reference workload of five concurrent AI coding sessions with thirty-six MCP servers. Absolute values depend on the client, model, and specific MCP set in use; the relative shape is stable. Verify for your deployment.
All figures measured on the reference setup described above. Process reduction, RAM reduction, and token savings depend on your MCP count, session count, model context window, and tool schema sizes. Results on other configurations may differ.
The reference implementation is distributed on PyPI and npm. Source under AGPL v3. The artifact runs locally; configured upstream servers and connectors may use the network. It is intended to be forked, studied, and critiqued.
Part of the Qualixar research platform. Maintained by Varun Pratap Bhardwaj.