ns-analyze-tracing

Analyzes N|Solid and OpenTelemetry traces to diagnose latency, HTTP errors, and service dependencies.

6|7|Updated Jun 4, 2026
One-click install
npx skills add https://github.com/nodesource/nsolid-plugin --skill ns-analyze-tracing-nodesource
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ns-analyze-tracing
Source: https://github.com/nodesource/nsolid-plugin/tree/main/skills/ns-analyze-tracing
Command: npx skills add https://github.com/nodesource/nsolid-plugin --skill ns-analyze-tracing-nodesource

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Diagnosing slow endpoints, API timeouts, and cascading failures across distributed Node.js services requires sifting through trace data manually. This Skill structures that investigation by analyzing supplied trace trees or querying live N|Solid tracing data to pinpoint bottlenecks and failing endpoints. ## Core Features & Use Cases - Trace Evidence Analysis: Parses provided trace IDs, span trees, status codes, and durations as authoritative evidence before querying live systems. - Live Tracing Queries: Uses N|Solid MCP tools to find slow requests via duration filters and failing endpoints via HTTP status code filters, with app-coverage reconciliation to avoid false absence claims. - Architectural Recommendations: Proposes fixes such as Redis caching, parallelizing independent requests with Promise.all, or message queues, and validates improvements after deployment. - Use Case: A user reports that a checkout API intermittently takes over 2 seconds. The Skill filters traces by duration, identifies the slow service and endpoint, explains the likely cause, and produces a markdown report with a validation plan. ## Quick Start Analyze the attached trace export and tell me which endpoint is causing the latency spike in my checkout service.

Frequently Asked Questions about ns-analyze-tracing

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

FAQPage Schema
How do I find slow endpoints in Node.js distributed traces?▼

Filter tracing data by duration using pipe syntax such as durations="1000|5000" to isolate spans between 1 and 5 seconds. The Skill returns a collapsed trace list showing the slow service, endpoint, status, duration, and trace ID for triage.

How do I analyze a trace ID from OpenTelemetry or N|Solid?▼

Provide the trace ID directly in your prompt and the Skill queries tracing with span_traceId for list-level rows. If you supply a full trace tree or export, it maps the parent-child span hierarchy directly without live queries.

Can I analyze traces without connecting to a live N|Solid console?▼

Yes. Supplied trace trees, exports, or span lists are treated as authoritative evidence and analyzed directly. Live tracing queries and connected-service discovery only run when no sufficient host-provided trace data exists.

Why does the tracing summary show no traces for my app?▼

The default tracing response is a paginated sample of 25 traces, so an absent app group does not prove the app lacks tracing. The Skill re-queries per app with the original filters, then once unfiltered, before any absence claim is justified.

What are the limitations of trace waterfall analysis with this approach?▼

The tracing MCP tool exposes only collapsed trace-list rows, not full waterfall details, so parent-child span claims require a user-supplied trace tree. Without that hierarchy, only service, endpoint, status, and duration evidence can be asserted.