skills.rest
eron1703
eron1703COMMUNITY
Explore all skills by this author →
GitHub

worker-stuck-protocol

Diagnose stuck worker agents and guide recovery via STOP, DIAGNOSE, REPORT workflow.

1|Updated Feb 6, 2026
Search Tags:#workflow#debugging#blockers#reporting#guardrails#stuck
One-click install
npx skills add https://github.com/eron1703/claude-config-loader --skill worker-stuck-protocol
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: worker-stuck-protocol
Source: https://github.com/eron1703/claude-config-loader/tree/main/skills/worker-stuck-protocol
Command: npx skills add https://github.com/eron1703/claude-config-loader --skill worker-stuck-protocol

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This protocol provides a structured, safe approach to recover from stall situations in worker agents, reducing wasted time and unsafe exploratory actions.

Core Features & Use Cases

  • STOP immediately when a stall is detected to prevent further actions
  • Diagnose whether the issue is a missing capability vs a real blocker
  • Report and escalate using the BLOCKED status and NEED_CAPABILITY / QUESTIONS guidance
  • Provides a clear, repeatable workflow that can be taught and audited

Quick Start

When a stall is detected, execute STOP, diagnose the cause, and report with a BLOCKED status to resume with supervisor guidance.

Frequently Asked Questions about worker-stuck-protocol

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

FAQPage Schema
How do I handle a stuck worker agent in Claude Code?▼

To handle a stuck worker agent, you need a deterministic STOP, DIAGNOSE, and REPORT workflow that halts progress, identifies missing capabilities or permission errors, and escalates blockers safely.

What is the best way to recover a worker agent stalled by permission errors?▼

The best way to recover a stalled worker agent is to execute a STOP action immediately, diagnose whether the permission error is a real blocker or a missing capability, and report it using a structured BLOCKED format.

How does the BLOCKED reporting format work for escalating workflow blockers?▼

The BLOCKED reporting format works by capturing specific blockers and escalating them with NEED_CAPABILITY or QUESTIONS guidance, ensuring supervisor intervention can resume the stalled worker agent.

When do I need a specific protocol for diagnosing worker stalls?▼

You need a specific protocol for diagnosing worker stalls when progress halts due to service outages or missing capabilities, preventing unsafe exploratory actions and wasted time during agent recovery.

Can I use this workflow to distinguish between a real blocker and a missing capability?▼

Yes, you can use the DIAGNOSE step within the workflow to distinguish between a real blocker, such as a service outage, and a missing capability, ensuring accurate escalation and recovery guidance.

Related Skills

Explore More Software Engineering Skills→
tcdenttcdent

agent-identity

Manage agent identities in Claude Code with automatic session renaming.

Community
Basic
nibojniboj
1

twelve-factor-concurrency

Define explicit web and worker process types for scalable concurrency.

Community
Intermediate
ruvnetruvnet
67.5k

flow-nexus-swarm

Automates cloud-based AI swarm deployments with event-driven workflows and multi-topology orchestration.

Community
Advanced
davidabeyerdavidabeyer
13

supervisor

Monitor blocked AI workers and respond to their pending questions.

Official
Basic