trace

Orchestrate competing tracer hypotheses to explain observed Claude team-mode results.

Updated Apr 11, 2026
One-click install
npx skills add https://github.com/Pryma-Tech/iris --skill trace-pryma-tech
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: trace
Source: https://github.com/Pryma-Tech/iris/tree/main/cli/skills/trace
Command: npx skills add https://github.com/Pryma-Tech/iris --skill trace-pryma-tech

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Orchestrates competing tracer hypotheses to explain observed results in Claude built-in team mode, enabling structured reasoning and faster root-cause discovery.

Core Features & Use Cases

  • Restates the observation, generates diverse hypotheses, and gathers evidence in parallel.
  • Spawns 3 tracer lanes by default for robust comparison and prioritization.
  • Produces a ranked synthesis with a critical unknown and a discriminating probe to guide next steps.
  • Suitable for runtime bugs, performance issues, architecture decisions, and config/orchestration questions.

Quick Start

Describe the observed result and start a team-mode trace with three distinct hypotheses.

Frequently Asked Questions about trace

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

FAQPage Schema
How does evidence-driven tracing help debug runtime anomalies in Claude team mode?

Evidence-driven tracing in Claude team mode orchestrates competing hypotheses to explain observed runtime anomalies. It gathers evidence in parallel across multiple tracer lanes, ranks the hypotheses, and produces a structured synthesis to accelerate root-cause discovery.

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

Tracing performance issues with competing hypotheses involves spawning distinct tracer lanes to gather evidence simultaneously. The process ranks each hypothesis based on gathered evidence, applies rebuttal rounds, and outputs a final synthesis with a discriminating probe for next steps.

Can I use structured tracing for architecture decisions and configuration behaviors?

Structured tracing supports architecture decisions and configuration behaviors by restating the observation and generating diverse hypotheses. It evaluates evidence across three lanes and delivers a ranked synthesis with a critical unknown to guide decision making.

How do I start a team-mode trace for a runtime bug?

Start a team-mode trace for a runtime bug by describing the observed result. The tracer generates three distinct hypotheses by default, gathers evidence in parallel, and ranks them to produce a structured final synthesis suitable for decision making.

Why does generating multiple hypotheses improve root-cause discovery for bugs?

Generating multiple hypotheses improves root-cause discovery by comparing diverse explanations of observed bugs. The three-lane tracing approach gathers evidence in parallel, applies rebuttal rounds, and prioritizes hypotheses to reveal the most likely cause.

Does Claude team mode tracing work without external dependencies?

Claude team mode tracing operates without external dependencies. It leverages the built-in team mode to orchestrate tracer lanes, gather evidence, and generate a ranked synthesis with a discriminating probe entirely within the workflow.