skills.rest
PMKhai
PMKhaiCOMMUNITY
Explore all skills by this author →
GitHub

orchestrator

Coordinate multi-agent workflows through staged architect, researcher, coder, reviewer, and debugger tasks.

Updated Apr 28, 2026
Search Tags:#automation#workflow#architecture#multi-agent#review#software-development#orchestrate
One-click install
npx skills add https://github.com/PMKhai/claude-config --skill orchestrator-pmkhai
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: orchestrator
Source: https://github.com/PMKhai/claude-config/tree/main/skills/orchestrator
Command: npx skills add https://github.com/PMKhai/claude-config --skill orchestrator-pmkhai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Multi-agent orchestration of complex workflows by coordinating Architect, Researcher, Coder, Reviewer, and Debugger agents in sequence.

Core Features & Use Cases

  • Stage-based orchestration: spawn Architect and Researcher in parallel, then route tasks to backend/frontend/full-stack coders, followed by a Reviewer and optional Debugger.
  • Deterministic task flow with a shared filesystem state at tasks/[project]/[task-id]/ to store SPECs, code summaries, and approvals.
  • End-to-end QA loop including review, issue handling, and optional Learner insights after approval.

Quick Start

Spawn the Architect and Researcher to initialize a task, then route coders based on the SPEC and observe the resulting review.

Frequently Asked Questions about orchestrator

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

FAQPage Schema
How do I coordinate multi-agent workflows for complex software development?▼

Multi-agent workflow coordination is handled by staging specialized agents—architects, researchers, coders, reviewers, and debuggers—in sequence with gate-driven stages to ensure traceable, auditable outcomes.

Can I run architect and researcher agents in parallel during task orchestration?▼

Yes, task orchestration supports optional parallelization, allowing you to spawn the Architect and Researcher agents simultaneously to initialize tasks before routing subsequent work to coders.

How does multi-agent orchestration track task state across review and debugging stages?▼

Multi-agent orchestration tracks state using a shared tasks filesystem, storing SPECs, code summaries, and approvals at tasks/[project]/[task-id]/ to maintain deterministic task flow across stages.

What is the best way to structure end-to-end QA loops in an automated software workflow?▼

End-to-end QA loops are structured by routing completed code to a Reviewer agent, followed by issue handling and an optional Debugger, with optional Learner insights generated after final approval.

Does multi-agent orchestration require external dependencies to manage staged handoffs?▼

No external dependencies are required to manage staged handoffs, as the orchestration enforces gate-driven stages and deterministic task spawning entirely through its internal design and shared filesystem.

Related Skills

Explore More Software Engineering Skills→
HillBombCreationsHillBombCreations

cmd-research

Automate bug research and tracking in the Vivreal stack.

Official
Advanced
vZulinvZulin

tweetclaw

Automate Twitter/X searches, posts, follower exports, and monitoring via OpenClaw.

Community
Advanced
NabilThangeNabilThange
1

lint-and-validate

Automate linting and static analysis for Node.js, Python, and TypeScript code.

Community
Intermediate
matheushenriquefsmatheushenriquefs

mentat-implement

Execute mentat plans slice-by-slice with TDD and gate checks.

Community
Advanced