trace

Generate competing hypotheses and parallel lanes for observation-based tracing.

Updated Apr 20, 2026
One-click install
npx skills add https://github.com/jimmi2051/oh-my-copilot --skill trace-jimmi2051
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: trace
Source: https://github.com/jimmi2051/oh-my-copilot/tree/main/plugins/omc-copilot/skills/trace
Command: npx skills add https://github.com/jimmi2051/oh-my-copilot --skill trace-jimmi2051

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ambiguity in observations often requires comparing multiple explanations; this skill coordinates structured tracing to surface and compare competing hypotheses across parallel lanes.

Core Features & Use Cases

  • Restate the observation and generate diverse hypotheses
  • Spawn a 3-lane tracer setup by default in Claude built-in team mode
  • Collect evidence for and against each lane, rank hypotheses, and produce a final synthesis with a critical unknown and a discriminating probe

Quick Start

Describe the observation and let the trace skill generate three tracer lanes, gather evidence for and against each lane, and output a ranked explanation.

Frequently Asked Questions about trace

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

FAQPage Schema
How do I perform root-cause diagnosis when multiple competing explanations could explain an observation?

Root-cause diagnosis for ambiguous observations is structured by generating competing hypotheses across parallel tracing lanes, collecting evidence for and against each, and ranking them to produce a final synthesis with a discriminating probe.

What is the best way to investigate runtime performance issues using competing hypotheses?

Investigating runtime performance issues works by spawning a 3-lane tracer setup in Claude team mode to evaluate diverse explanations, rank their strength based on collected evidence, and output a recommended discriminating probe.

Can I use team mode to coordinate parallel evidence collection for architecture questions?

Yes, team mode coordinates parallel evidence collection for architecture questions by spawning three tracer lanes that gather supporting and refuting evidence, rank hypotheses by strength, and synthesize the critical unknown.

How do I start a structured trace to evaluate multiple hypotheses for a system failure?

To start a structured trace, describe the observation and let the skill generate three default tracer lanes in team mode, which then gather evidence for and against each lane and output a ranked final explanation.

Does structured hypothesis tracing work without external dependencies or components?

Structured hypothesis tracing works without external dependencies or components, relying entirely on Claude's built-in team mode to coordinate lanes, collect evidence, and synthesize the critical unknown for the observation described.

Why should I use parallel tracing lanes instead of a single linear investigation?

Parallel tracing lanes prevent premature commitment to one explanation by forcing evidence collection for and against multiple competing hypotheses simultaneously, ultimately surfacing the critical unknown and a discriminating probe to resolve ambiguity.