Pulse MCP Server

Bring Pulse traffic intelligence directly into your AI workflows.

We're introducing the Pulse MCP server, a Model Context Protocol interface for querying observed network traffic, requests, issues, PII signals, and aggregate metrics from Pulse-backed data.

Released: Feb 20, 2025

circle-check

What's New

MCP Access to Core Pulse Resources

AI assistants can query structured data across:

  • Connections

  • Requests

  • Traffic (aggregates, availability, bandwidth, performance)

  • Issues (policy and API errors)

  • PII detections

Rich Tooling for Fast Investigation

The server exposes count/list/get/histogram/frequency patterns plus specialized tools for:

  • Request duration and tag analytics

  • Traffic summaries, time-series, location and geo breakdowns

  • Issue/error exploration

  • PII summaries and distributions

Filter-Driven Analysis

Use field-based and fuzzy filters (endpointId, vendorId, path, status, direction, etc.) to narrow investigations quickly.

Time-Aware Querying

Built-in support for RFC3339 timestamps and configurable granularity (minute to year) makes trend analysis straightforward.


Why It Matters

Pulse MCP turns raw traffic intelligence into conversational, automatable workflows for engineering, security, and operations teams:

  • Ask natural-language questions about live or recent traffic

  • Investigate incidents with reproducible query patterns

  • Build AI assistants that reason over real production telemetry

  • Reduce manual dashboard hopping during outages


Example Outcomes

  • "Which endpoints had the highest p99 latency in the last hour?"

  • "Show issue frequency by error for this vendor in prod"

  • "Count requests with auth failures and list affected paths"

  • "Summarize detected PII types by endpoint"


Feedback

Have ideas for new MCP tools or filters?

Last updated