dos-unstick

Analyze recurring blocked runs and propose structural fixes by cause.

18|4|Updated Jun 10, 2026
One-click install
npx skills add https://github.com/anthony-chaudhary/dos-kernel --skill dos-unstick
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dos-unstick
Source: https://github.com/anthony-chaudhary/dos-kernel/tree/main/claude-plugin/skills/dos-unstick
Command: npx skills add https://github.com/anthony-chaudhary/dos-kernel --skill dos-unstick

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps teams diagnose repeated BLOCKED, DRAIN, and STALLED runs so they can stop the underlying cause instead of manually clearing the same failure over and over.

Core Features & Use Cases

  • Mines recent run archives to capture blocker text, run id, iteration, and stall cost.
  • Normalizes blocker language to canonical cause keys from the workspace reason vocabulary.
  • Clusters recurring causes by frequency and stall cost, then proposes one structural fix per cluster for operator review.
  • Use it when an agent fleet keeps stalling on the same issues and you need a systemic remediation plan rather than a one-off unblock.

Quick Start

Use the dos-unstick skill to analyze the latest blocked runs in this workspace and return the highest-priority recurring causes with one structural fix each.

Frequently Asked Questions about dos-unstick

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

FAQPage Schema
How do I find the root cause of recurring blocked runs in a DOS workspace?

To find the root cause of recurring blocked runs, you analyze sequential run archives and blocker text to cluster repeated failures by frequency and stall cost. This surfaces structural causes keeping agent fleets stalled instead of requiring manual unblocks.

What is the best way to stop agent fleets from repeatedly stalling on the same issues?

The best way to stop recurring stalls is to normalize blocker language to canonical cause keys and cluster them by stall cost, then propose one structural remediation per cluster. This provides a systemic remediation plan rather than clearing one-off failures.

How do I cluster BLOCKED and DRAIN outcomes by stall cost in a run archive?

You cluster BLOCKED and DRAIN outcomes by reading sequential run records from the workspace and normalizing the blocker text to a declared reason vocabulary. The system then groups these recurring causes by frequency and stall cost for review.

Do I need a declared reason vocabulary to analyze recurring blocker issues?

Yes, a declared reason vocabulary is required to normalize blocker language into canonical cause keys. The analysis applies to DOS workspaces storing run archives alongside this vocabulary to accurately cluster structural causes.

Can I route proposed structural fixes directly to dos decisions without modifying code?

Yes, you can route proposed structural fixes directly to dos decisions add without modifying code. The analysis reads workspace paths and canonical causes to generate remediation proposals for operator review.