Analytics for Windsurf

Windsurf by Codeium is designed around flow state — Cascade AI works alongside you as a persistent collaborator. BetterMeter plugs into Windsurf via MCP so you can track both how Cascade uses your MCP tools and query your site analytics without breaking flow.

Add BetterMeter to Windsurf

Windsurf supports MCP servers through its configuration system. Add BetterMeter and Cascade can access all your analytics tools natively within the IDE.

{
  "mcpServers": {
    "bettermeter": {
      "command": "npx",
      "args": ["-y", "@bettermeter/cli@latest", "mcp"],
      "env": {
        "BETTERMETER_API_KEY": "bm_..."
      }
    }
  }
}

Track Cascade's tool invocations

If you build MCP servers that Windsurf connects to, BetterMeter tracks every tool call Cascade makes. Windsurf's Cascade operates differently from other AI clients — it maintains context across turns and tends to chain tool calls more aggressively. Understanding these patterns helps you optimize your tool design.

Flow-state analytics queries

Windsurf's philosophy is keeping you in flow. With BetterMeter connected, you can ask Cascade about your analytics mid-task. Deploying a new feature? Ask about traffic changes. Investigating a bug report? Check which pages are affected. The data surfaces where you already work.

• 01 Cascade integration — Cascade calls BetterMeter tools as part of multi-step workflows
• 02 Persistent context — Analytics data carries across conversation turns
• 03 Inline answers — Traffic stats appear right in the editor pane
• 04 Multi-site aware — Switch between projects and their analytics seamlessly

Cascade vs other AI clients

BetterMeter identifies Windsurf/Cascade in MCP client metadata, letting you compare its usage patterns against Claude Code and Cursor. See which AI client invokes tools most efficiently, which has higher error rates, and how invocation patterns differ. This is especially valuable if your MCP server serves multiple AI IDEs.