watchdog

Detect and classify stalled automation in Open Harness workflows.

28|3|Updated Mar 27, 2026
One-click install
npx skills add https://github.com/mifunedev/openharness --skill watchdog-mifunedev
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: watchdog
Source: https://github.com/mifunedev/openharness/tree/main/.claude/skills/watchdog
Command: npx skills add https://github.com/mifunedev/openharness --skill watchdog-mifunedev

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Stuck or slow automation loops across Open Harness workflows waste time and block progress; this watchdog detects stalls and moves work forward safely.

Core Features & Use Cases

  • Monitors cron/autopilot loops, draft PR workflows, and tmux sessions to distinguish slow progress from real stalls.
  • Applies safe completion actions to move automation forward without unsafe side effects.
  • Use cases include auto-finalizing stale draft PRs, restarting stalled sessions, and preserving workspace integrity during long-running automation.

Quick Start

Configure the watchdog with default settings to begin monitoring and automatically flag or fix stalled automation.

Frequently Asked Questions about watchdog

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

FAQPage Schema
How do I detect stalled automation in CI-CD workflows?

You can detect stalled automation by applying objective signals and terminal verification states to monitor cron-based autopilot loops, classifying slow progress versus real stalls to trigger safe completion actions.

What is the best way to auto-finalize stale draft PRs?

Auto-finalizing stale draft PRs is handled by monitoring draft PR workflows to identify stalls and applying safe mutation rules that move work forward without unsafe side effects.

How does tmux session monitoring prevent stuck automation loops?

Tmux session monitoring prevents stuck automation by tracking active sessions across a repository, detecting when processes stall, and triggering safe completion actions to preserve workspace integrity.

Can I monitor automation stalls across multiple repositories?

Monitoring automation stalls applies to one repository by default, but you can use optional --repo overrides to extend watchdog coverage to additional repositories as needed.

How do safe mutation rules prevent unsafe automated changes?

Safe mutation rules prevent unsafe automated changes by applying objective signals and terminal verification states, ensuring completion actions only execute when automation stalls are genuinely confirmed.

When should I avoid using automated stall detection for git workflows?

Automated stall detection should be avoided when workflows lack clear terminal verification states, as applying safe completion actions without objective signals risks triggering unsafe mutations in active git processes.