network-performance

Query Datadog NPM connections, DNS traffic, and NDM device inventory via the pup CLI.

5|2|Updated Mar 9, 2026
One-click install
npx skills add https://github.com/lauhon/pi --skill network-performance-lauhon
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: network-performance
Source: https://github.com/lauhon/pi/tree/main/skills/network-performance
Command: npx skills add https://github.com/lauhon/pi --skill network-performance-lauhon

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Investigating network issues across services and infrastructure requires manually querying Datadog Network Performance Monitoring and Network Device Monitoring APIs, which is slow and error-prone without knowing the right commands, dimensions, and metrics. ## Core Features & Use Cases - Connection Analytics: Query aggregated network flows grouped by client/server service, host, zone, or port, with TCP metrics like retransmits, timeouts, resets, and RTT latency. - DNS Traffic Analysis: Analyze DNS query volumes, response times, and failures grouped by domain or record type. - Device Monitoring: List, filter, and inspect network devices (routers, switches, firewalls), view interface status, and manage device tags. - Use Case: When users report slow API responses, query connections grouped by client and server service, identify flows with high rtt_micro_seconds or tcp_retransmits, and pinpoint the degraded network path. ## Quick Start Ask the agent to show network connections in production for the last hour grouped by client and server service.

Frequently Asked Questions about network-performance

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I query network connections in Datadog NPM?

Run pup network connections with --from and --to time parameters, optionally adding --group-by for dimensions like client_service and server_service. Use --tags to filter by environment or service and --limit to control result size.

How to analyze DNS traffic with Datadog network monitoring?

Use pup network dns with a time range and group by network.dns_query to see top domains by query volume. You can also group by network.dns_record_type or filter by client_service tags to investigate specific services.

What credentials does Datadog network monitoring require?

The pup CLI requires three environment variables: DD_API_KEY for your Datadog API key, DD_APP_KEY for your application key, and DD_SITE for your Datadog site, which defaults to datadoghq.com.

How do I find network devices that are down in Datadog NDM?

Run pup network devices list with --filter-tag="status:down" to show only non-responsive devices. You can also filter by warning status, location, or device type, then use pup network devices get for details.

What do high TCP retransmits indicate in network monitoring?

High tcp_retransmits values indicate packet loss or network congestion between client and server. Combined with tcp_timeouts and elevated rtt_micro_seconds, they suggest network quality issues on that connection path.

What is the maximum result limit for Datadog network queries?

The maximum limit for network connection queries is 7500 results, with a default of 100. If you exceed the limit, reduce the limit parameter or add more specific tag filters to narrow the query.