skills.rest
mabdulrauf
mabdulraufCOMMUNITY
Explore all skills by this author →
GitHub

agent-tracing

Record per-step execution snapshots and inspect traces via a Node.js CLI.

Updated Apr 6, 2026
Search Tags:#debugging#cli#agent-tracing#context-engine#execution-snapshots#llm-calls
One-click install
npx skills add https://github.com/mabdulrauf/wazone-back --skill agent-tracing-mabdulrauf
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-tracing
Source: https://github.com/mabdulrauf/wazone-back/tree/main/.agents/skills/agent-tracing
Command: npx skills add https://github.com/mabdulrauf/wazone-back --skill agent-tracing-mabdulrauf

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Agent tracing provides a transparent, debugging view into agent execution by recording step-by-step snapshots and context engine data, enabling developers to diagnose prompts, tool usage, and memory injections.

Core Features & Use Cases

  • Records per-step ExecutionSnapshot data to disk for later inspection.
  • Exposes a CLI to list, inspect, and view traces, steps, messages, tools, and memory.
  • Captures context engine input and output events, including model, knowledge, tools, and user memory.
  • Supports partial (in-progress) snapshots and finalization to complete traces.
  • Provides a terminal viewer to render traces and step details.

Quick Start

Run the agent-tracing CLI from the repository root to view the latest trace and begin debugging.

Frequently Asked Questions about agent-tracing

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

FAQPage Schema
How do I debug LLM agent execution step by step?▼

Debug LLM agent execution by recording per-step snapshots to disk that capture LLM calls, context engine data, and tool usage. A CLI viewer lets you inspect these traces to diagnose prompts and memory injections.

What is agent tracing and when do I need it for debugging?▼

Agent tracing records transparent execution snapshots and context engine data for debugging. You need it when diagnosing agent prompts, tool usage, or user memory injections across multiple execution steps.

How do I inspect agent context engine data and tool usage from the CLI?▼

Inspect context engine data and tool usage by running the agent-tracing CLI from your repository root. It exposes commands to list, view, and inspect traces, steps, messages, tools, and memory.

Can I view in-progress agent execution snapshots before a trace finalizes?▼

Yes, you can view in-progress agent execution snapshots. The system supports partial snapshots written to the .agent-tracing directory during execution, which are finalized to complete traces afterward.

Does this agent debugging tool require specific dependencies to run?▼

No specific dependencies are required. The agent debugging CLI is implemented as a Node.js or Bun-based tool that writes snapshot data directly to the .agent-tracing directory for inspection.

What's the best way to audit user memory injections during agent execution?▼

Audit user memory injections by capturing context engine input and output events in execution snapshots. The CLI viewer renders these memory injections alongside tool usage and LLM calls for detailed inspection.

Related Skills

Explore More Software Engineering Skills→
alceupassosalceupassos

agent-tracing

Record and inspect AI agent execution snapshots with a CLI tool.

Community
Intermediate
duwenjiduwenji

agent-tracing

Inspect JSON execution snapshots and step-level traces from .agent-tracing.

Community
Advanced
kubonykubony
10

gcp-snapshot

Automate create, restore, list, and cleanup of GCP disk snapshots via gcloud CLI.

Community
Advanced
kingmacthkingmacth

agent-tracing

Record agent execution snapshots to disk and inspect traces via CLI.

Community
Advanced