agent-worker-specialist

Executes assigned swarm tasks while reporting progress through claude-flow memory coordination.

70.1k|8.4k|Updated Jun 2, 2025
One-click install
npx skills add https://github.com/ruvnet/claude-flow --skill agent-worker-specialist
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-worker-specialist
Source: https://github.com/ruvnet/claude-flow/tree/main/.agents/skills/agent-worker-specialist
Command: npx skills add https://github.com/ruvnet/claude-flow --skill agent-worker-specialist

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinating task execution across a multi-agent swarm requires constant status visibility, and without a disciplined worker protocol, tasks get duplicated, blockers go unreported, and results are lost. This Skill defines a worker agent that executes assigned tasks while continuously reporting state through shared memory.

Core Features & Use Cases

  • Structured Status Reporting: Stores task-received, progress, blocked, and completion states in the claude-flow coordination namespace using memory_usage calls.
  • Specialized Work Modes: Provides templates for code implementation, analysis, and testing workers that share results with the swarm.
  • Dependency and Blocker Handling: Checks shared dependencies before starting work and reports blocking conditions immediately.
  • Use Case: In a hive-mind coding swarm, a queen-coordinator assigns a feature implementation; the worker reports progress every step, shares created files and test results, and delivers a completion record with performance metrics.

Quick Start

Invoke the worker-specialist agent to execute an assigned task and report its progress through swarm memory coordination.

Frequently Asked Questions about agent-worker-specialist

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

FAQPage Schema
How do I coordinate worker agents in a claude-flow swarm?

Assign each worker a unique ID and have it store status, progress, and completion records in the coordination namespace via mcp__claude-flow__memory_usage calls. The queen-coordinator reads these keys to track task state across the swarm.

How does a worker agent report task progress?

The worker stores JSON progress objects under keys like swarm$worker-[ID]$progress, including steps completed, current step, percentage, blockers, and files modified. Updates are recommended every 30-60 seconds during execution.

What happens when a worker's dependencies are not ready?

The worker checks the shared dependencies key before starting. If dependencies are missing, it stores a blocked record listing what it is waiting for instead of beginning work, preventing wasted execution.

Can multiple workers collaborate on the same task?

Yes, the parallel collaboration pattern lets peer workers divide work based on capabilities, sync progress through shared memory, and merge results on completion. Workers should not start unassigned work or make autonomous decisions.

What are the limitations of memory-based swarm coordination?

Coordination depends on all agents consistently reading and writing the agreed memory keys, so malformed keys or missed updates can desynchronize the swarm. It also requires the claude-flow MCP memory tool to be available in the environment.