Skip to main content

Pulse MCP Tools

Tools available in the Pulse MCP server for performance monitoring.

pulse_overview

Get a performance overview.

Parameters

NameTypeRequiredDescription
sincestringNoTime range (default: 1h)

Examples

“How’s our app performing?”
pulse_overview(since: "1h")
Returns:
  • Apdex score
  • Average response time
  • P95/P99 latencies
  • Throughput (requests/min)
  • Error rate

pulse_slow_requests

Find slow requests.

Parameters

NameTypeRequiredDescription
threshold_msnumberNoDuration threshold (default: 1000)
sincestringNoTime range
limitnumberNoMax results

Examples

“What requests are taking over 2 seconds?“
pulse_slow_requests(threshold_ms: 2000, since: "24h")

pulse_throughput

Get request throughput over time.

Parameters

NameTypeRequiredDescription
sincestringNoTime range (default: 1h)
intervalstringNoGroup by: minute, hour, day

Examples

“What’s our throughput today?“
pulse_throughput(since: "24h", interval: "hour")

pulse_errors

Get requests that resulted in errors.

Parameters

NameTypeRequiredDescription
sincestringNoTime range
limitnumberNoMax results

Examples

“Show me failed requests from the last hour”
pulse_errors(since: "1h")

pulse_endpoints

Get performance breakdown by endpoint.

Parameters

NameTypeRequiredDescription
sincestringNoTime range (default: 1h)
sortstringNoSort by: slow, frequent, errors

Examples

“Which endpoints are slowest?”
pulse_endpoints(since: "24h", sort: "slow")
“Which endpoints get the most traffic?“
pulse_endpoints(since: "24h", sort: "frequent")

pulse_trace

Get details of a specific trace.

Parameters

NameTypeRequiredDescription
trace_idstringYesThe trace ID

Examples

“Show me trace trace_abc123”
pulse_trace(trace_id: "trace_abc123")
Returns the full trace with all spans (database queries, HTTP calls, etc.)

pulse_metrics

Get custom metrics.

Parameters

NameTypeRequiredDescription
namestringYesMetric name
sincestringNoTime range

Examples

“What’s our queue size metric?”
pulse_metrics(name: "queue_size", since: "1h")