Windsurf via Tokonomix

Add the Tokonomix MCP server to Windsurf's Cascade agent — cross-family consensus as a tool, right inside the flow.

MCP server (Cascade)AI-native IDEmacOS · Windows · Linux

Windsurf — AI-native IDE with Cascade, via Tokonomix

What is Windsurf?

Windsurf (by Codeium) is an AI-native IDE whose Cascade agent plans and edits across your codebase. Cascade supports the Model Context Protocol (MCP), which is the clean way to bring Tokonomix in — as a consensus tool the agent can call.

Why MCP, not a base-URL swap

Windsurf runs its own model layer, so the integration that always works is MCP: add the Tokonomix server and Cascade gains a tokonomix_consensus_ask tool, regardless of which chat model Windsurf itself uses.

Why add Tokonomix to Windsurf?

Cascade does autonomous multi-file work — a consensus check on a risky step catches what one model misses.
One key exposes tokonomix_consensus_ask, tokonomix_single_ask, tokonomix_list_models as tools.
Consensus is metered per call in your Tokonomix dashboard — use it only where it earns its keep.
EU / sovereignty routing applies to every consensus call the agent makes.

Add the Tokonomix MCP server

Add the Tokonomix MCP server in Windsurf’s MCP configuration (Settings → Cascade → MCP, or the mcp_config.json), then reload so the tools appear to Cascade.

{
  "mcpServers": {
    "tokonomix": {
      "command": "npx",
      "args": ["-y", "tokonomix-mcp"],
      "env": { "TOKONOMIX_API_KEY": "tok_live_your-key" }
    }
  }
}

Or just ask Cascade

Tell Cascade: "Add the Tokonomix MCP server — npx -y tokonomix-mcp — to my MCP config with my key." Then reload Windsurf so the new tools register.

Verify it works

After reload, ask Cascade to "get a consensus verdict" on a change.
Confirm the result comes back as the MCP tool result.
Check the metered call in your Tokonomix dashboard.

Model selection

Which model and endpoint should my agent use?

Pick by what your agent does, then swap the model id any time — every model in the catalog works through the same key.

What your agent does	Use this
Edits files / calls tools and streams responses (most coding agents)	An Anthropic model via the Anthropic endpoint (tools + streaming both work)
Calls tools but does not need streaming (any model)	The OpenAI-compatible endpoint with a single model id
Wants tools and streaming on a non-Anthropic model	Run non-streaming for now — streaming tool-calls on other providers is not enabled yet
Plain chat, no tools	Either endpoint; streaming is fine
Wants a cross-model verdict (consensus) on a high-stakes answer	Set model to tokonomix-consensus on the OpenAI-compatible endpoint — note: consensus rejects tool requests (no tool-calling)

Endpoint & API reference (technical)

anthropic-messages → https://tokonomix.ai/api/anthropic — tools ✓, streaming ✓ (use an Anthropic model). openai-completions → https://tokonomix.ai/api/v1 — tools ✓ (single-model passthrough), streaming with tools not yet enabled. Plain chat: streaming works on both. Consensus: openai-completions, model tokonomix-consensus, no tools. Swap the model id to any slug returned by GET /api/v1/models.

Browse all models →