MCP Server Accuracy and Tools Enhanced

You can now access more accurate and complete data when interacting with Everflow through the Model Context Protocol (MCP) server. This update resolves several critical data gaps, ensuring AI agents no longer receive silent errors or incomplete tracking and reporting metadata.

What's New

Previously, certain MCP queries returned incomplete click details or incorrect conversion event statuses, forcing a manual fallback to the REST API. This release populates missing geographic and device fields for clicks, ensures conversion events correctly return their event IDs, and exposes custom metrics in performance reports. Additionally, we have introduced new standalone tools for global search, pixel-fire logs, and network-wide cap pacing.

Key Capabilities

• Accurate conversion event tracking: Resolves an issue where conversion event IDs were returned as zero, ensuring AI agents can correctly identify and track secondary conversion events.
• Enriched click metadata: Populates region, city, browser, platform, device type, OS version, and user IP directly within click search results.
• Custom metrics integration: Surfaces custom metrics and formulas in performance reports, allowing agents to retrieve and display tailored network metrics.
• New standalone tools: Adds dedicated tools for searching global terms, viewing pixel-fire logs, monitoring the on-hold conversion queue, and running network-wide cap reports.
• Improved API reliability: Fixes a conversion paging bug that caused duplicate or skipped rows, and resolves several server errors on malformed reporting requests.