trace

Display chronological agent flow traces across hooks, keywords, skills, agents, and tools.

1|Updated Mar 17, 2025
One-click install
npx skills add https://github.com/ozby/node-pubsub --skill trace-ozby
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: trace
Source: https://github.com/ozby/node-pubsub/tree/main/.codex/skills/trace
Command: npx skills add https://github.com/ozby/node-pubsub --skill trace-ozby

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Displays the runtime trace of how hooks, keywords, skills, agents, and tools interact during a session, helping you diagnose flow issues and performance bottlenecks.

Core Features & Use Cases

  • Timeline view: see the chronological sequence of events across a session, including hooks, skills, agents, and tools.
  • Summary view: aggregated counts and metrics such as hook fires, keywords detected, skills activated, mode transitions, and tool performance.
  • Filtering & focus: narrow the trace by event types (hooks, skills, agents, keywords, tools, modes) and by session scope (latest or specific sessions).
  • Use case: debugging complex agent workflows and optimizing response times.

Quick Start

Run trace_timeline with no arguments to display the latest session timeline, then run trace_summary to review aggregated metrics.

Frequently Asked Questions about trace

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

FAQPage Schema
How do I visualize an agent workflow trace to debug performance bottlenecks?

To visualize an agent workflow trace, use the trace_timeline tool to display a chronological sequence of hook, skill, agent, and tool events. This timeline view helps you pinpoint flow issues and identify performance bottlenecks across a live session.

What is an agent trace summary and how does it aggregate workflow metrics?

An agent trace summary provides aggregated metrics for a session, including counts for hook fires, keywords detected, skills activated, and tool performance. It is generated using the trace_summary tool to review overall workflow behavior at a glance.

Can I filter agent flow traces by specific event types like hooks and tools?

Yes, you can filter agent flow traces by specific event types. The trace functionality allows you to narrow the timeline view down to hooks, skills, agents, keywords, tools, and modes, letting you focus only on the events relevant to your debugging scope.

How do I get the runtime trace for a specific agent session instead of just the latest one?

To get the runtime trace for a specific agent session, you can adjust the session scope filter. While running trace_timeline with no arguments defaults to the latest session, specifying a session ID or scope narrows the chronological trace to that exact session.

What's the best way to diagnose complex agent flow issues during a live session?

The best way to diagnose complex agent flow issues is to run trace_timeline to view the chronological event sequence, then run trace_summary to review aggregated metrics. Combining the timeline and summary views reveals how interactions cause response delays.