What does the Agent Loop Observer do?

It provides a unified view of all agent activities across different repositories, showing which loops are currently running, which have finished, and which have crashed.

Can I attach to a running agent loop session?

Yes, the skill automatically generates and displays the specific tmux attach commands needed to jump directly into an agent's live terminal session.

Does it save historical data for completed loops?

Yes, it pulls data from git-committed history files to show previous configurations, prompts, iteration snapshots, and summaries even after the process finishes.

How does it detect crashed loops?

The skill compares start/end registry events against active tmux sessions; if a start event exists without an end event and the tmux session is dead, it identifies the loop as crashed.

Agent Loop Observer

Name: Agent Loop Observer
Author: JSai23

byJSai23

•

分析与监控

Monitors and manages autonomous agent loops across multiple repositories with real-time status tracking and crash detection.

Agent Loop Observer provides comprehensive visibility into autonomous agent activities across your entire development environment. By aggregating data from cross-repo registries, per-repo tracking logs, and git-committed history, it allows users to track running, completed, and crashed agent loops without manual file searching. It offers detailed execution timelines, iteration snapshots, and immediate terminal access via tmux integration, making it an essential tool for developers orchestrating multiple AI-driven workflows simultaneously.

主要功能

011 GitHub stars

02Detailed iteration timelines including durations and exit codes

03Automated crash detection and state recovery for failed processes

04Cross-repository visibility into all active and historical agent loops

05Comprehensive loop history access from git-committed records

06Integrated tmux session management for direct terminal attachment

使用场景

01Identifying and cleaning up ghost agent processes after a system crash or unexpected termination

02Reviewing historical performance and stop reasons for long-running autonomous agent tasks

03Tracking the progress of background agents running in multiple project directories simultaneously

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jsai23/claude-tooling loop-loop-observe-a

For use in Claude.ai and ChatGPT

主要功能

011 GitHub stars

02Detailed iteration timelines including durations and exit codes

03Automated crash detection and state recovery for failed processes

04Cross-repository visibility into all active and historical agent loops

05Comprehensive loop history access from git-committed records

06Integrated tmux session management for direct terminal attachment

使用场景

01Identifying and cleaning up ghost agent processes after a system crash or unexpected termination

02Reviewing historical performance and stop reasons for long-running autonomous agent tasks

03Tracking the progress of background agents running in multiple project directories simultaneously

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jsai23/claude-tooling loop-loop-observe-a

For use in Claude.ai and ChatGPT