skills.rest
Derek-X-Wang
Derek-X-WangCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

team-harness

Coordinate parallel Claude Code agent teams for feature development and debugging.

Updated Mar 21, 2026
Search Tags:#workflow#debugging#orchestration#Claude-code#agent-teams#team-harness#parallel-dev
One-click install
npx skills add https://github.com/Derek-X-Wang/skills --skill team-harness
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: team-harness
Source: https://github.com/Derek-X-Wang/skills/tree/main/team-harness
Command: npx skills add https://github.com/Derek-X-Wang/skills --skill team-harness

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Kick off a Claude Code agent team with specialized teammates that work in parallel and communicate directly. Unlike feat-harness (sequential subagents), team-harness uses agent teams where teammates see errors, fix them, test, and loop — without waiting for a central orchestrator.

Core Features & Use Cases

  • Sets up a three-teammate workflow: Engineer, Dev Watcher, and QA Tester, each owning non-overlapping files and responsibilities.
  • Provides pre-flight checks to ensure the agent teams are enabled and ready.
  • Includes a standard coordination pattern: environment startup, feature development, testing, and iterative bug fixing.
  • Supports debugging and feature development workflows where parallel execution accelerates feedback loops.

Quick Start

Spawn a three-teammate team (engineer, dev watcher, QA tester) to start the feature or debugging task.

Frequently Asked Questions about team-harness

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

FAQPage Schema
How do I coordinate parallel Claude Code agent teams for feature development?▼

To coordinate parallel Claude Code agent teams, you can use a harness that spawns an Engineer, Dev Watcher, and QA Tester to work simultaneously. This setup shares context directly to identify and fix issues quickly without waiting for a central orchestrator.

What is the difference between sequential subagents and parallel agent teams?▼

Sequential subagents wait for a central orchestrator, whereas parallel agent teams communicate directly and loop independently. Teammates see errors, fix them, and test in real-time, accelerating feedback loops during feature development and debugging.

How do I prevent file conflicts when running parallel AI agents?▼

Prevent file conflicts by enforcing non-overlapping file ownership for each parallel agent. The workflow assigns specific responsibilities to the Engineer, Dev Watcher, and QA Tester, keeping work scoped, predictable, and conflict-free.

Does team-harness require pre-flight checks before spawning agent teams?▼

Yes, pre-flight checks are required to ensure parallel agent teams are enabled and ready. This validation step occurs before spawning the Engineer, Dev Watcher, and QA Tester to execute the structured 5-6 task plan per teammate.

Can I use parallel agent teams for debugging workflows?▼

Yes, you can use parallel agent teams for debugging workflows. The Engineer, Dev Watcher, and QA Tester work simultaneously, sharing context to identify issues, fix errors, and test iteratively to accelerate bug resolution.

Related Skills

Explore More Software Engineering Skills→
maxwell2732maxwell2732
39

session-init

Verify git state, Plans.md status, and harness-mem resume context before session tasks.

Community
Intermediate
alinaqialinaqi
705

iterative-development

Refine code through automated test-driven development loops with completion criteria.

Community
Advanced
discountedcookiediscountedcookie

subagent-workflow

Coordinate subagent orchestration with session tracking and structured handoffs.

Community
Advanced
tmalcolm-0607tmalcolm-0607

mad-parallel

Orchestrate wave-based parallel feature implementation across isolated git worktrees.

Community
Advanced