Observability

Stream PAI multi-agent activity to a WebSocket-enabled dashboard.

17.4k|2.3k|Updated Sep 8, 2025
One-click install
npx skills add https://github.com/danielmiessler/Personal_AI_Infrastructure --skill observability-danielmiessler
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Observability
Source: https://github.com/danielmiessler/Personal_AI_Infrastructure/tree/main/.claude/Skills/Observability
Command: npx skills add https://github.com/danielmiessler/Personal_AI_Infrastructure --skill observability-danielmiessler

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires bun, vue, typescript, and includes scripts (resource) components.

What problem does it solve?

This Skill eliminates the frustration of not knowing what your AI agents are doing, giving you complete visibility into complex workflows so you can identify bottlenecks and optimize performance.

Core Features & Use Cases

  • Real-Time Dashboard: Live WebSocket streaming of all agent activity across sessions.
  • Debug Workflows: Quickly identify stuck agents, looping processes, or failed tool calls.
  • Use Case: When your multi-agent system seems unresponsive, activate this Skill to see exactly which agent is blocked and why.

Quick Start

Simply say "start observability" to launch the monitoring dashboard and server, then watch as all agent tool calls, session starts/ends, and hook events display in real-time.

Frequently Asked Questions about Observability

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

FAQPage Schema
How do I debug multi-agent workflows in real-time?

Real-time debugging of multi-agent workflows involves monitoring agent activity, tool calls, and session events as they execute. This Skill provides a WebSocket-enabled dashboard that streams live agent activity, letting you identify stuck agents, failed tool calls, and bottlenecks instantly without waiting for logs to be written.

Can I monitor agent sessions and tool calls live?

Yes, this Skill monitors agent sessions, tool calls, hook events, and timelines in real-time via WebSocket streaming. It watches JSONL history files, broadcasts live updates to a Vue dashboard, and provides complete visibility into multi-agent activity without requiring persistence or manual log parsing.

What's the best way to set up real-time observability for multi-agent systems?

Set up observability by running a Bun + Express server backend paired with a Vue + TypeScript dashboard frontend. The server streams agent activity through WebSocket connections, file-watches history logs, and broadcasts events in-memory, giving you instant visibility into workflow execution without configuration overhead.

Do I need a separate logging system for agent observability?

No. This Skill provides in-memory streaming observability without requiring external persistence. It watches your existing JSONL history files, streams events live to the dashboard, and applies optional event summarization hooks—eliminating the need for separate logging infrastructure.

Why would my multi-agent system seem unresponsive?

Multi-agent systems can stall due to blocked agents, failed tool calls, or looping processes. Observability dashboards reveal which agent is unresponsive and why by showing live tool calls, session events, and hook execution—letting you diagnose and fix bottlenecks immediately.