orchestrate

Coordinate multidisciplinary agents to execute product development plans from specs.

Updated Mar 5, 2026
One-click install
npx skills add https://github.com/jaisonerick/spec-plugin --skill orchestrate-jaisonerick
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: orchestrate
Source: https://github.com/jaisonerick/spec-plugin/tree/main/plugins/spec-plugin/skills/orchestrate
Command: npx skills add https://github.com/jaisonerick/spec-plugin --skill orchestrate-jaisonerick

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Managing a multidisciplinary product development workflow is complex, error-prone, and requires continuous coordination across architecture, story breakdown, implementation, design, and QA; this Skill centralizes orchestration so plans progress reliably from spec to verified code.

Core Features & Use Cases

  • Plan Selection & Resume: Inspect roadmap and plan specs, determine existing state, and let the user choose to resume or restart.
  • Agent Orchestration: Spawn and manage specialized agents (architect, product-owner, engineer, designer, qa) with isolated worktrees, merge flows, and unique task lifecycles.
  • Parallel Execution & QA Looping: Analyze story dependencies for safe parallelism, run QA spec creation in parallel, coordinate sequential QA runs, and manage automated fix cycles with environment failure handling.
  • Progress & Retrospective: Maintain PROGRESS.md, produce summaries, run retrospectives, and clean up team resources.

Quick Start

Use the orchestrate skill to run a full plan by telling the assistant which plan to run from specs/roadmap.md and confirming how to proceed.

Frequently Asked Questions about orchestrate

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

FAQPage Schema
How do I coordinate multiple agents to execute a product development plan end-to-end?

Multi-agent orchestration coordinates specialized agents like architect, product-owner, engineer, designer, and qa to execute a product development plan from spec to verified code. It manages isolated worktrees, task lifecycles, branching, merging, and QA cycles across the team.

What is the best way to manage parallel execution and QA loops across development agents?

Parallel execution and QA looping analyzes story dependencies for safe parallelism, runs QA spec creation concurrently, coordinates sequential QA runs, and manages automated fix cycles with environment failure handling to ensure reliable progress.

Do I need agent definitions and spec files to use multi-agent orchestration for product development?

Yes, multi-agent orchestration requires agent definitions in .claude/agents/, task and team APIs, worktree setup hooks, and spec files under specs/ to manage branching, merging, and QA cycles for plans defined in specs/roadmap.md.

Can I resume an existing product development plan if it was interrupted midway through implementation?

Plan selection and resume inspects the roadmap and plan specs, determines the existing state, and lets the user choose to resume or restart the workflow. Progress is tracked in PROGRESS.md to maintain continuity across sessions.

How does multi-agent orchestration handle retrospectives and progress tracking for product development?

Multi-agent orchestration maintains PROGRESS.md, produces summaries, runs retrospectives, and cleans up team resources after plan completion. This centralizes coordination so plans progress reliably from spec to verified code.

What are the limitations of using isolated worktrees for managing parallel agent tasks?

Isolated worktrees require worktree setup hooks and task and team APIs to manage branching and merging correctly. If environment failures occur during automated fix cycles, the orchestration system handles them, but complex story dependencies may limit safe parallelism.