wide-events

Design wide-event schemas with high-dimensional span attributes and tail sampling.

Updated Mar 12, 2026
One-click install
npx skills add https://github.com/mpsuesser/workspace --skill wide-events-mpsuesser
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: wide-events
Source: https://github.com/mpsuesser/workspace/tree/main/dotconfig/pi/agent/skills/wide-events
Command: npx skills add https://github.com/mpsuesser/workspace --skill wide-events-mpsuesser

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Wide events solve the problem of observability logs being optimized for writing instead of querying, making incident debugging slow when one request spans many services and systems.

Core Features & Use Cases

  • Single canonical log line per request: Consolidate request, user, business, performance, and outcome context into one enriched event to make correlation easy.
  • High-dimensional, queryable fields: Design span/log attributes so structured queries can answer questions like “which failures affected premium users under a specific feature rollout.”
  • Tail-focused sampling strategy: Preserve errors, slow requests, and VIP users while sampling the rest to control observability cost without losing critical signal.

Use case examples include designing instrumentation coverage for distributed services, reviewing which span annotations to capture, and validating that incident questions are answerable from the emitted fields.

Quick Start

Ask the AI to help you design a wide-event schema for a trace span by listing the identity, user, business, performance, and outcome attributes you should attach to that span for your main request flow.

Frequently Asked Questions about wide-events

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

FAQPage Schema
What is a wide event in distributed systems observability?

A wide event is a single canonical log line per request that consolidates identity, user, business, performance, and outcome context into one enriched event for distributed systems observability. This high-dimensional model makes cross-service correlation easy during incident debugging.

How do I design structured logging attributes for incident debugging?

To design structured logging for incident debugging, attach high-dimensional queryable fields to trace spans covering identity, user, business, performance, and outcome attributes. This ensures emitted telemetry can answer incident-driven questions like which failures affected premium users under a specific feature rollout.

What's the best way to sample OpenTelemetry spans without losing critical signals?

The best way to sample OpenTelemetry spans without losing critical signals is applying a tail sampling policy that retains errors, p99-slow requests, and VIP users while sampling the rest. This controls observability costs while preserving the telemetry needed for incident debugging.

Can I use wide events to correlate logs across multiple microservices?

Yes, you can use wide events to correlate logs across multiple microservices by consolidating fragmented, low-context logs into one comprehensive event per request. This high-dimensional model with structured span annotations replaces scattered logs optimized for writing with queryable events optimized for correlation.

Why does incident debugging take so long with fragmented observability logs?

Incident debugging takes long with fragmented observability logs because they are optimized for writing instead of querying, making it hard to correlate context across distributed services. Wide events solve this by consolidating request, user, business, performance, and outcome data into one enriched event.